From bus stops and billboards above rows of shuttered shops in freezing January wind, the faces of intubated coronavirus victims stare down at us.

This is the British government’s bleak new lockdown ad campaign; huge posters with the heavy saturation and high contrast redolent of cheap horror film publicity which demand that you look suffering people in the eyes and justify selfish rule-breaking. It employs the schlocky attack ad style pioneered by American political strategists, except the subject of the attack is not an opposition candidate, it is you. Such ads can be brutally effective in elections, but they display a shocking lack of understanding of what can actually persuade people to shift their behavior.

Most simply, a behavior change message must ask people to do something they are capable of. One of these ads reads, “Look him in the eyes and tell him you can’t work from home,” as if the crowds on London’s morning rush hour Tube trains are furtively sneaking into their workplaces for a thrill. In the first lockdown, one-third of workers were compelled to continue working. Over the summer, the government was even more assiduous than employers in ordering workers into the office, telling people flatly to return to work or lose their jobs. This appeared to be a major market intervention on behalf of the beleaguered landlords of chain cafés.

Secondly, most communications professionals will caution against shouting at people. In general life, people are rarely persuaded of anything through guilt-tripping and hectoring. In mass communication, the role of fear and guilt in an effective message is somewhat more nuanced and weighted by broader sociological and material factors, but the principle generally holds. The high-handed style of attack, shame-and-blame, assuming guilt, and appealing to authority is advised against in persuasion from conspiracy theories to substance abuse. The government’s own Behavioural Insights Team tends to advocate gentler nudge-based approaches to persuasion.

The now-classic “don’t think of an elephant” communications exercise demonstrates that raising a subject focuses attention primarily on the subject itself — not whatever you are trying to say about it. This has profound adverse consequences when the subject is lockdown compliance. We are social beings who emulate the behavior of others around us; if our peers are following the rules, we will (particularly when the rules are avoiding contact with said peers.) If people believe rule-breaking is common enough to merit ever-present grim billboards, it reduces the cost and danger of rule-breaking. “If everyone is breaking the rules, why should I bother following them?” This also increases mistrust, which may encourage people to dob in their neighbors to a chaotic and overstretched policing regime, and poses huge risks to the sense of shared sacrifice and collective spirit we need to weather this crisis.

Most critically, government communications in a crisis should be grounded in reality. The reality is that rule-breaking is not statistically significant. Public polling shows overwhelming support for restrictions, and more detailed analysis shows 85-90 percent levels of compliance. Unprecedented controls have been implemented with remarkably high levels of consent. The most strident opponents of infection control have not been recalcitrant citizens, but establishment commentators often in ideological lockstep with the government. A large section of the public in both March and January moved faster than the government to limit social contact, and yet the government’s high polling numbers reflect the considerable goodwill it possesses for whatever measures it adopts — even after presiding over an enormous, tragic, and global record casualty count.

This level of public tolerance is remarkable in the face of government breaking another golden rule of crisis communications — clarity. From June to December, we pivoted through a dizzying array of alert levels, tiers, partial lockdowns, and subsidized virus-spreading opportunities. At the beginning of this month, we went from “schools are perfectly safe” to “leave the house and you will kill people” within twenty-four hours. The only serious loss of goodwill came with the government’s refusal to hold Dominic Cummings to account.

This British Medical Journal blog details the extent to which misrepresenting the scale of noncompliance has been a sustained government strategy. Tailored communications carefully targeted toward those few who are more likely to break the rules — not a general ad campaign — would be a sensible tool in the arsenal. But a mass campaign based in shaming, discipline, and control that is utterly unjustified by reality is both inappropriate and deeply unimaginative. Lack of imagination, of course, is a more charitable interpretation. The arguments above are not revolutionary, and the government is surrounded by communications and behavior change experts. Widespread perceptions of rule-breaking do achieve one thing; they cause people to blame each other rather than a government that has stubbornly refused to follow successful countries in using the time bought by partial lockdowns to pursue eliminating the virus.

Most messaging problems are rooted in politics. The “stay at home, protect the NHS, save lives” mantra is a clear and useful slogan. Bolstering its effectiveness requires providing support to follow the rules, learning from more effective measures like the furlough scheme. The billboard posters could tell us how to access support with daily tasks or mental and physical well-being. They could be telling us that statutory sick pay had been brought into line with every comparable country to aid self-isolation. They could allow you to report your employer forcing you into work, rather than your neighbor going for a second jog. This would of course require imagination beyond poster design, enough to confront the short-termist, profit-seeking incentive structures that have irreparably damaged the pandemic response both in Britain and beyond.