If I could offer one piece of advice to a profession fighting its greatest battle in recent memory, I think it would be this: please don’t open a second front on domestic politics. It’s not going to work.

Your newfound status as caped heroes battling the coronavirus will not survive contact with the culture wars, which suck everyone down into the mud and where the good guys are indistinguishable from everyone else.

All you will achieve is the politicisation of your profession and the associated suspicion of your recommendations, rendering you useless in the domestic struggle and weakened in the battle against the ongoing crisis.

Quite why so many public health experts felt the need to burn up their accumulated trust and prestige in order to defend mass gatherings on political grounds is therefore a mystery to me. It’s as though a collective amnesia has gripped the profession. People who were one day warning against the dangers of attending the funerals of family members, and of the unconscionable risks involved in visiting our friends, have pivoted cleanly to saying that there is no convincing argument against gathering tens of thousands in close proximity.

What makes this worse is the utter transparency of the reasoning involved; the pivot is entirely partisan. Dr Tom Frieden, the former Director of the US Centre for Disease Control, tells us that the threat posed by the protests is “tiny” compared to the threat of the government acting “in ways that lose community trust”. Just a few days before he was emphasising the deadly risk of social gatherings.

Greg Gonsalves, an Assistant Professor at Yale, took a different tack: protesters are “balancing competing risks to their communities”, and voting with their feet on which is the greatest. It is entirely consistent to argue “for social distancing in February, March, April, and May”, and now to argue “for exceptions for social gatherings such as protests”. Anyway, we’re past the peak and people are wearing masks.

The grab bag of arguments on offer indicates that there really isn’t a single clear and compelling reason to believe that protesters screaming “I can’t breathe” won’t spread a disease that kills its victims by filling their lungs with fluid. Instead, there are strong partisan reasons to avoid pointing this out; people want to be good allies to the black community, and this means letting the disease spread faster there. Black lives matter, but not as much as political approval.

Saying that protesters have “voted with their feet” when weighing these risks is a non-starter; the public health profession is out there telling the world that the risks are lower. Claiming to follow revealed preference while reinforcing that decision is clearly inconsistent, and it is far from clear that the necessary analysis to determine which “risk” is greater has been performed.

A rationale for protests that follows this utilitarian calculation needs harder backing than the feelings of the man on the street. If you have good evidence that the number of quality-adjusted life years lost from protesting will be less than the number lost by failing to protest, then you need to present it, and you need to include in that analysis the effect on compliance with lockdown overall.

As Frieden was saying, maintaining trust in the proportionality and efficacy of your measures is critical in generating compliance. If people can look outside and see a crowd of thousands spreading the disease, what exactly do they gain from following the rules themselves? If enforcement is absent or selective, it’s hard to form reasonable expectations around what is required, and what will happen if the rules are broken.

The UK has just had precisely this conversation at length around the behaviour of senior adviser to the Prime Minister, Dominic Cummings, who was found to have driven to Durham while potentially infected with coronavirus to isolate on his family farm.

Cue a week of furious debate as to whether this was technically permitted within the rules or a flagrant violation, during which figures advising the government on lockdown compliance opined that letting Cummings off had “trashed all the advice we have given on how to build trust and secure adherence”. A few weeks later and the same advisers are tweeting their support for the protests in America.

It’s an interesting question. Will a single high profile individual getting away with leaving the capital encourage more people to break the rules than several thousand people congregating in the heart of London? My gut instinct says “no”, but I could be wrong on this. What I do not believe is that the first could “trash” all advice on securing compliance while the second has no effect.

The second reason we should disregard the argument that people have “voted with their feet” is that the entire point of having binding rules is to prevent them from doing so. It is entirely possible for breaking lockdown to be rational on an individual level while being undesirable from a social perspective; the protests put at risk people who did not attend by bringing the virus back to their communities.

The idea that we can simply follow the decisions of the protesters in assessing risk is one without credibility when the consequences of their decisions may fall on others. The public health profession knows this; it is fundamental to their entire approach to handling communicable disease. So why are they setting it aside now?

I am clearly not alone in finding the sudden change in attitudes confusing. Jeffrey Flier, former dean of Harvard Medical School, told Politico that “the sudden change in views of the danger of mass gatherings has been disorientating, and I suspect it has been for many Americans”.

But perhaps we’re simply taking too simplistic a view of the matter. Perhaps the public health profession really has weighed the damage to its reputation and the willingness of people to believe its advice against the benefits of interceding. After all, they’ve voted with their feet.

This article was originally published on Sam Ashworth-Hayes’s blog, Marginally Productive.