One intriguing aspect of the scientific discussion about #COVID19 measures is how entirely unscientific it feels to me. "Letting young people get infected by the virus is immoral" or "lockdowns are evil" are not scientific statements.
1/

Science can inform on the current situation (e.g. number of undetected #COVID19 cases) and provide plausible projections under different interventions, or absence thereof. Though, science does generally not, in itself, inform on the morality of any such intervention.
2/

Thus, when a scientist expresses their views on what #COVID19 measures the population should adopt, they do not tend to follow the science or go against it, but make a subjective, emotional statement about what they believe is morally/ethically right or wrong.
3/

There is a scientific approach to decide on the merit of #COVID19 measures, based on 'scientific utilitarianism'. Scientists can estimate the cost of measures in terms of gain/loss of lives, years of life, or more complex indices including e.g. wellbeing and education.
4/

A utilitarian approach to #COVID19 may seem appealing to some, but it is not without its challenges. While the calculations are based on science, the choice of which metric to maximise/minimise is not. This becomes particularly difficult when qualitative factors are included.
5/

For example, a #COVID19 model minimising deaths in the short term is likely to give a very different answer about what public health measures perform 'best', relative to one maximising years of life over the next decades and even more so from one including wellbeing/education.
6/

The choice of the metric to minimise/maximise (e.g. deaths, years of life or anything more inclusive of general health) falls outside the realms of science. It is a moral/ethical problem that should not be left to decide by scientists alone.
7/

It feels somewhat unlikely to me that a societal consensus could emerge about what metric #COVID19 public health measures should aim to minimise/maximise (e.g. deaths or years of life), over what timescale and at what geographical scale (country level or more globally).
8/

In the unlikely event a societal consensus emerged on what metric to minimise/maximise, one problem remains. Long-term projections from complex mathematical models tend to be imperfect, because the models are flawed, poorly parameterised or the situation changed unexpectedly.
9/

That said, despite the limitations of an utilitarian approach to #COVID19, injecting a bit more science in the scientific debate on #COVID19 should be beneficial to everyone.
10/