(RE)Constructing the missing COVID-19 UK impact assessment for 2020/21


Given any new proposal for government policy in the UK, the appropriate departmental Civil Servants develop a document called an IA (variously known as an ‘Impact Analysis’ or ‘Impact Assessment’). This document considers, in the widest sense, all the costs and benefits to the United Kingdom of the proposed policy.  In particular, it attempts to calculate the ‘social value’ of the policy via an agreed-cross Government methodology overseen by HM Treasury.  Social value is measured in UK pounds but is not simply a financial estimate but includes, for example, reductions or gains in the quality of life of those affected and deaths either caused or prevented. The use of pounds as the unit of social value is conventional for, as we shall see in the following, a pound of social value may be more or less than a ‘pound in your pocket’.

The IA, including both the estimation of social value and any unquantifiable costs and benefits, is presented to Ministers. Before the policy can go forward, Ministers are expected to sign the IA to say they believe that the benefits outweigh the costs. This may either be because the quantifiable social value is positive and the unquantifiable benefits are not significant or because the unquantifiable benefits clearly (at least in the mind of the Government) outweigh any loss of quantifiable social value.  In other words, to formally declare that, in their view, the policy is economically ‘rational’.  If these conditions are not met, Civil Servants may ask for a formal direction to proceed with a policy which is economically ‘irrational’ – the responsibility (say in front of the Public Accounts Committee) now belonging entirely to the Minister.

Of course, in an emergency it is recognised that developing an IA is impractical and matters must proceed on the best judgement of those responsible for the situation. However, as an emergency begins to extend to weeks and months and the policy implications are seen to be substantial and long term, the development of an IA would be expected from the lead Government department.

As someone who in their time in advising Government had produced and quality assured IA’s, I was surprised that no IA supporting the extreme suppression, or  as it became known, ‘lockdown’ policy was made public – especially as previous pandemic policy making had rejected and argued strongly against  the ideas of  ‘lockdown’ type policies.

In March 2020, I therefore requested under the Freedom of Information Act a copy of the relevant IA.  Late in 2020 I received a reply indicating no such IA existed. Since this time, the UK Covid Inquiry has also established that no such official IA was prepared at any point. 

I thus constructed my own, which I finished in January 2021, following Government Guidance both from the viewpoint of autumn 2020 and from that of March 2020 when the suppression (lockdown) policy was first initiated. 

As I suspected from my knowledge of previous discussions of pandemic policy, the result was clear – the suppression or lockdown policy was ‘irrational’.  From the perspective of March 2020, using the most favourable possible assumptions for the lockdown policy and very conservative assumptions for the old mitigation policy, the quantified social value was essentially the same with all unquantifiable benefits on the side of mitigation compared to suppression. For any more reasonable set of assumptions, both the quantifiable social value and the unquantifiable benefits were substantially on the side of mitigation.  This was even clearer from the autumn 2020 perspective. Further, given the lack of information available to me as a member of the public, I had given the benefit of the doubt to many of the lockdown policy assumptions. I believe that an official IA with even a little better information, and benefiting from Governmental resources to process that information, would surely have produced an even clearer view still, of the ‘irrationality’ of the lockdown policy.

What follows (in the attachment) is the impact assessment I completed.   It is a document of its time. Much of the detail has been superseded but this has little impact on the conclusions, especially on the picture as it would have been seen before the decision to move to the suppression policy via ‘lockdown’.  One point does require some discussion: there is a section on the likelihood of long term natural and vaccine immunity (due to declining personal immunity or the development of new strains). This discussion in hindsight is clearly overly optimistic. Despite this, the conclusion of the discussion was conservative, favouring the lockdown policy i.e. that whatever immunity was developed it was only sufficient to get through to vaccine introduction – and that the vaccine would be introduced whether or not natural immunity existed.  The over-optimism does not therefore impact the analysis.  

It is worth emphasising that the IA follows UK Government practice on how decisions should be taken as justified, in particular in its strict ‘utilitarian’ calculation of ‘social value’ (though mitigated by the consideration of unquantifiable benefits).  I have always been a critic of this methodology but it does represent how the UK Government would have tackled the problem if it had been working ‘by the book’.  In any case, it is a reasonable starting point for a more nuanced discussion taking account of other factors, and importantly recording them for future reflection.

It has also been objected that the value of a human life is in fact, infinite or at least immeasurable and that any measure of ‘social value’ is thus meaningless.  That is to neglect the fact that any entity may have a whole range of different, incommensurable, values attached to it.  I would agree that from a religious point of view, or in the context of one’s family, every life is of immeasurable value, however we are here interested in allocative value – how to best use limited public resources.  At its starkest, if we have to choose between two actions, otherwise equal, but one which will inevitably lead to one death, and another that equally inevitably will lead to two, which do we choose?  The loss of social value due to the lockdown policy has led to fractured lives, and inevitably illness and deaths. If the mitigation strategy had been followed the resources would have been potentially available to improve the quality of life for many and save more lives.

Before the IA proper there is a comparison of the lockdown policy with previous Government policy with similar health impacts indicating why the questions put in the IA are important.

The IA has had a limited circulation amongst those arguing against the lockdown policy. I have sent it twice to the UK Inquiry and received polite replies but no further interest. Hence, I submit it here as a historical document with the aim of recording the type of analysis that was available to Government during the early days and months of the Covid pandemic.

About the Author

Until his retirement, 6 years ago, Dr Peter Grove chaired the Department of Health’s Senior Analytical and Economic Review Committee advising Ministers on the effectiveness and cost effectiveness of new Departmental Programmes.  He was also the lead analyst on Vaccination and Immunisation and Emergency (including Pandemic) Preparedness.  His role included chairing SPI-M the Government’s pandemic modelling committee and other similar groups.  His background includes work in Theoretical Physics, Artificial Intelligence, Criminology, Statistics and Operational Research.

PDF version

Loader Loading…
EAD Logo Taking too long?

Reload Reload document
| Open Open in new tab

Download [1.32 MB]