A few years ago a campaign ran in Britain that was meant to get children to eat more vegetables, and it did something that must have struck every nutrition officer as wrong at first. It did not tell the children that vegetables were healthy. It said nothing about vitamins, about concentration in class, about life later on. Instead it declared the vegetables to be the enemy. The carrots, the peas, the broccoli became villains from the earth that had to be defeated by being eaten. It was called Eat Them to Defeat Them, and in the evaluation 45 percent of children and 31 percent of parents who had seen the advertising reported that more vegetables were eaten afterwards.
The interesting thing here is not the success. The interesting thing is that the very same message, wrapped in the honest, reasonable, nutritionally correct argument, would reliably have worked worse, and measurably so. A Stanford study by Turnwald, Boles and Crum in 2017 offered the same vegetables in a cafeteria over an entire term, sometimes under health-focused, sometimes under indulgent labels, and under the indulgent label 41 percent more people helped themselves. It was literally the same carrot.
Anyone who lays those two observations side by side already has in front of them the most uncomfortable property of almost all well-meant communication: the correct argument and the effective message are rarely the same thing, and surprisingly often they are opponents. And this is exactly where the problem of German political communication begins, which in four of its biggest fields, pensions, climate, migration and defence, is almost always entered through the danger.
The reflex that everything begins with
A politician talking about pensions naturally assumes that the topic is important enough to justify the audience's attention. That assumption feels so obvious that it is never voiced, and it is wrong all the same, not because pensions are unimportant but because importance and attention have nothing whatsoever to do with each other. The person you want to reach in the moment of the broadcast is thinking about their fridge, about an argument from yesterday, about what is due tomorrow morning, and about a dozen other things that are more immediate to their life than the demographic curve of the statutory pension system.
This is neither stupidity nor indifference in the moral sense, but the ordinary economy of attention, which is always scarcer than the volume of topics laying claim to it. The default state of an audience toward almost any political topic is not rejection, not hostility, but simply absence, the fact that the topic has so far supplied not a single reason why it should be relevant to this particular person right now, today.
Anyone who wants to break through that state, and both politics and media want to, reaches almost automatically for the one tool that works fast enough: danger. A threat produces attention because the brain has an evolutionarily favoured processing route for danger signals that pushes other content aside for a moment. This works, in the narrow sense that the attention does in fact arise. What happens afterwards lies beyond the measurement range of those who produced the attention, and so it occurs to no one that the next step, in which the awareness created is supposed to turn into some kind of action, has been damaged by the very mechanism that produced the attention in the first place.
Why people would rather look away
The reason for this has a name, even if it never comes up in German talk shows. The economist George Loewenstein first pinned it to an observation about stock investors: people check their portfolio more often when prices rise and stop looking when prices fall, even though the information would be exactly equally valuable in both directions. Karlsson, Loewenstein and Seppi called it the ostrich effect, after the bird that supposedly buries its head in the sand. At its core it is information avoidance under perceived loss of control: when information only produces pain without enabling action, it is entirely rational not to take it in.
Transferred to political communication this means: a threat communicated loudly enough to be perceived, yet offering no action that could show up in one's own life, produces not readiness to act but shielding. The person who changes nothing about their behavior after the third climate report in the evening programme is not irrational. They have learned that information about the climate makes no action accessible to them personally that would make a recognisable difference, and have consequently filed it as irrelevant to themselves. This is not a defect of the psyche, this is a psyche protecting itself against consequence-free discomfort.
Climate research has by now replicated this finding so often that in practice it really ought not to be ignored any longer. The Yale Program on Climate Change Communication shows consistently that threat framing on diffuse, distant, complex risks does not raise the willingness to act but raises psychological distance, precisely the gap that classifies the problem as real but not mine. Fear without a way to act does not move the topic closer, it pushes it further away.
The 2023 debate around the so-called heating law is the textbook case in real time. What was communicated was essentially an obligation with a deadline, combined with financial uncertainty whose scale stayed unclear for months, something even a later analysis by the Progressive Centre and Robert Habeck's own admission of communication mistakes record.
The result was not that people switched their heating. The result was a political mobilisation against the measure itself. The perceived threat was high enough to trigger resistance but too diffuse to enable constructive action, and that is the ostrich effect in its most destructive form: not looking away, but active counter-mobilisation.
The wall you build for yourself
Here is the point where the diagnosis turns from an annoying detail into a structural finding. For the short, fast win of breaking indifference with danger, you pay a price that only comes due later and then makes the actual task more expensive than if you had never started. You buy the attention of the first second by using that same second to erect a barrier against everything that follows. Anyone who starts with fear has already sabotaged taking the audience along before saying the first constructive sentence.
With that, the uncomfortable truth has been spoken that you simply have to let stand before talking about solutions: the system of mass media and the system of behavior change are structurally incompatible. Media did not learn to work with fear out of malice, they discovered that fear produces clicks because people react to fear, and built a working business model on it. That is not a flaw in their system, that is the clean optimisation toward their target variable, namely attention. It is only that this target variable is, at its core, irreconcilable with the societal target variable, actual behavior change.
And now comes the correction that makes the problem worse rather than easier. One could assume the participants know this on some level and are knowingly accepting the damage. That is probably too optimistic. More likely the line of thought is never drawn at all, because attention and behavior change are owned in completely separate silos, by different people, measured against different metrics. The editor optimises for reach. The campaign strategist optimises for attention points in the decisive week. The communication scholar publishes on framing effects in journals neither of the other two reads. And no one in this chain is responsible for what actually happens at the end, when viewers leave the heating on after the broadcast, keep postponing their pension planning, or next time put their cross where their anger was articulated loudest.
That the one cannibalises the next step is therefore not a failure of actors who ought to know better. It is the normal way a system works that was built for attention and never for anything else. Anyone who wants to change it must first recognise at all that the form of communication is itself part of the problem, and not a neutral channel through which better content could be sent.
What it looks like when someone actually means it
What the other path looks like in concrete terms can be shown through two more British campaigns, both of which forgo something that seems almost indispensable in German politics: the wagging finger and the danger.
The first is a series of spots by the state health authority from the late eighties, at a time when AIDS was spoken of almost exclusively in the register of catastrophe. The most notorious British spot of those years, Don't Die of Ignorance from 1987, showed a tombstone chiselled in stone and a rock bursting out of the dark, scored by the voice of John Hurt, and it is exactly what we now know about threat framing: maximum fear, maximum memory of the poster, minimal information about what the individual is actually supposed to do day to day. Set against it, a little later, was a poster campaign with an invented figure called Mrs. Dawson, a factory worker in a condom plant, whose message ran roughly to the effect that one should keep Mrs. Dawson nice and busy and use condoms. The point is not in the joke but in the tonality: when an older factory worker treats the subject as matter-of-factly as production figures, it loses all embarrassment and becomes something entirely ordinary that one simply talks about. The Mrs. Dawson piece is documented to this day in the collection of the Science Museum.
This is the third messenger category that is almost always overlooked in political debate. We know the expert, who works through authority, and we know the similar one, who works because they are like the audience. Mrs. Dawson is neither, she resembles no 25-year-old a creative team wanted to reach, and she works in a normalising way all the same, because the right question was never whether this person looks like the audience but what this person signals about the behavior. She signals: this is normal, there is nothing here to be alarmed by.
The second campaign is a long-running series for a British railway company, animated in the style of the children's book series around Enid Blyton's Famous Five. A group of children travels into the country by train while others take the same route by car, and the children on the train arrive first. Not a word about sustainability, not a word about CO2, no obligation. Instead concrete characters, a concrete decision and a concrete outcome that you can bring about yourself. That is exactly what the Zeitenwende speech of 2022 lacked completely, that attempt to reverse the German stance on defence in a single term: abstract historical weight, a diffuse threat a thousand kilometres away, and a number, a hundred billion in special funds, without any story for anyone to hold on to.
That almost nothing of Zeitenwende survives three years later as a term, while the societal stance on defence readiness has changed little structurally, is no surprise but memory research. The psychologist Ian Begg showed as early as 1972 that people recall concrete, vivid phrasing at 36 percent and abstract phrases at only 9 percent, a factor of four. A term like Zeitenwende is pure abstraction, and abstraction evaporates.
Why being allowed to act pulls harder than being made to act
What these working campaigns share is not the good cheer but that they give the person something to do that they decide themselves. The children defeat the vegetables, they are not obliged to eat. The Famous Five choose the train, they are not admonished to take the rail. Behind this apparent triviality sits one of the most robust findings in behavioral psychology: Ellen Langer showed in 1975 that people demand markedly more for a lottery ticket they chose themselves, and are reluctant to give it up, compared with an assigned ticket of identical odds. The feeling of having chosen creates a value no obligation, however urgent, can create.
Obligation framing takes exactly that away from the person. It tells them they must do something, and in the same breath strips them of the dignity of their own decision, which is why the predictable reflex is not cooperation but defiance. Communication that wants to change behavior instead plugs into a motivation that is already there, into the need for competence, for control, for belonging, and it does so without ever demanding that the person adopt the authority's perspective.
From this follows the starting point I call the Indifference Premise, and at first it sounds almost like a provocation: you begin with the assumption that the audience is entirely right not to care about the topic. Not as rhetorical modesty, but as an operating basis. Anyone who starts from this point, and not from the conviction that their own topic is important enough to deserve attention, arrives at a completely different question. No longer how to make the topic relevant, but into which already existing everyday decisions, identities and motives the desired behavior can be built, without the person even noticing that they are processing a political message.
The communication that changes behavior over the long run is structurally more invisible, harder to attribute and slower than the kind that produces attention. In a system whose only currency is visibility, it stands almost no chance.
This is the real asymmetry, and it explains why better knowledge does not solve the problem. The actor who would apply the Indifference Premise would have to give up, in the short term, the attention fear delivers instantly, against the long-term and within no electoral period measurable hope that normalised, agency-based communication eventually changes behavior. As long as visibility remains the only rewarded variable, none of the participants gains anything by taking the slower road. Anyone who wants to change that must work not on the message but on what the system rewards, and that is a considerably harder task than a better script.