As in previous cases (e.g. the foot and mouth outbreak of 2001) the tensions between scientific advice and policy decisions is difficult, even though the UK government is relatively pro-science and has established systems for channelling advice.
The government’s Scientific Advisory Group for Emergencies (SAGE) gives confidential advice to the government, following established civil service practice, but some details leak out. It was initially divided as to what strategy to pursue, with some advising, early on, against lock-down measures to slow the virus, but then changing its mind as the situation in Italy unfolded (which fitted in with the results from the Imperial/Ferguson modelling projections). This was reported on in BuzzFeed’s previous article “10 Days That Changed Britain: “Heated” Debate Between Scientists Forced Boris Johnson To Act On Coronavirus“.
The UK government, under pressure on its relatively late imposition of lock-down measures, is now taking the line that they are always “guided by the scientific and medical advice”. Whilst this seems to be, at least somewhat, true BuzzFeed now reports that ‘Scientists Advising The UK Government On The Coronavirus Fear Boris Johnson’s Team Is Using Them As “Human Shields”’. In other words, that some of the blame will be on the scientists rather than the politicians.
There is a dilemma here. If a complex set of models and evidence suggest some outcomes then policy makers are faced with a difficult choice, do they (a) trust that advice and base their decisions on it to some extent or (b) essentially take little notice of the advice and go with their own experience or politics. Either way they are open to blame if (in hindsight) the alternatives to their choices look better. Even scientists in the field find it hard to understand the models and the data, a policy actor has no chance of doing this so has to decide whether to trust the advice or not. If there is a long track record of the models, techniques, data this is relatively uncontroversial and the line between science and policy (weighing the costs against avoidance of risks) can be made, but when the science is emerging and the situation is complex with lots of uncertainties then this is much more difficult.
The problem, from a democratic point of view, is that:
- the whole process of advice (SAGE etc.) is hidden and not open to inspection,which is why we mostly have to rely on reporting such as BuzzFeed’s (in this case Katie J.M. Baker‘s)
- even if the advice process is open, sometime the modelling and data is not (the Imperial team have still not released the code of their simulation model which was described by Neil Fergusson in a tweet as “…(thousands of lines of undocumented C) [Iwritten] 13+ years ago to model flu pandemics…“)
The tension between advice based on complex science and decision-making is not going to be easily solved, but there are a series of measures that would help, namely:
- Established and open procedures that operate when scientists give advice to governments, including that any models and data on which this is based is made immediately open to critique and review by other researchers.
- That, at least, any caveats that the scientists add to advice is published, so that exactly what was said and what was later decided upon is clear. Then both scientists and decision-makers can be clearer about their own areas of responsibility.
- That journals, when publishing work like Fergusson’s flu epidemic model, have to document and make public their code, so that other experts have time to check and critique the work before a crisis hits and the model gets relied upon for advice. Good modelling takes time and a lot of eyes.
- That policy actors learn what such models, advice etc. can and can not be used for, so they are better informed about the limitations and advantages of such an approach (e.g. as in reading this Blackett report).
- That a distance is kept between scientists and policy actors, so that there is not so much pressure on the scientists to give any particular kind of advice, or frame their conclusions in any particular way (though financial incentives to pursue research in useful directions is acceptable).
- That the scientific advice process does not, in effect, exclude other stakeholders or limit the range of the political input that takes place concerning the decision.
(Some of these issues are discussed in this paper)