I was born male and I identify as male, but according to the calorie guide on a Tesco value cheesecake, I’m a family of four. Calorie counts on packaged supermarket food have been around long enough for the joke to be an old one. Despite that, somehow, their introduction has not made us slimmer.
We have just reached the fourth anniversary of calorie counts becoming mandatory on restaurant menus for companies with more than two hundred and fifty employees. In the scheme of things, this is only a small intrusion into our lives by well-intentioned public health policy. The intrusion is small and the effects dubious — but the approach is revealing.
When it was introduced, the virtues of the scheme were trumpeted. Boris Johnson’s government explained the rules “will help the public to make healthier choices when eating out”. Jo Churchill, then Public Health Minister, talked about the importance of this “building block in our strategy to support and encourage people in achieving and maintaining a healthier weight.” Entirely lacking from the 2021 announcement was any mention of how the impact of the regulations would be measured.
Obesity is a problem. People don’t like being fat, being fat is bad for them, and we are drifting into a world in which obesity is becoming the norm. The problem deserves being taken seriously. We should be open to new intrusions into our lives that make us healthier.
Yet good intentions alone are not enough. What matters — indeed, the only thing that matters — is whether such interventions achieve measurable, beneficial outcomes. In medicine we teach our students that every drug and every intervention has costs. The question is never whether a treatment is safe. The question is whether it does more good than harm.
With public health interventions, harms can be hard to measure. There has been no real effort to gauge the costs on businesses of producing calorie counts. That’s discreditable, but there are other costs too — ones which genuinely can’t be counted, but count all the same. Having our noses rubbed in the calorie counts of our meals is a minor but real diminishment of our lives, an intrusion into our daily business. Perfectly reasonable to accept that if it offers substantial benefits, insane to tolerate it if it doesn’t. The difference matters.
Measuring the impact of interventions like calorie counts is not hard, but it demands effort. Does calorie labelling affect what people eat? Cluster randomisation — bringing in regulations in some places and not others as part of a trial — can provide a reliable answer. Calorie labelling has been brought in across the developed world but no proper attempt to get such an answer, in any country, was ever made.
Demanding public health policies are held to a high standard should be common ground for libertarians and nanny-staters alike
A second choice for measuring their impact would be large scale observational data on which one could conduct an interrupted time series analysis — not simply looking at before and after figures, but adjusting them for underlying trends. No serious attempt at this was made either, not anywhere in the world. Read the evidence — collected here — and the conclusion that calorie counts are “potentially meaningful”. Examine the individual studies behind this verdict, and you’ll find they’re either poorly designed, disappointing in their results, or both. There is no way a drug would get licensed on the basis of evidence this bad. Yet governments, who are used to implementing policies whose effects they never properly evaluate, feel this is fine.
Calorie counts on menus were not brought in because science showed they were worth the trouble, but because the winds of public health fashion blew in their direction. Soon they may well be ushered out for the same reason: a recent flurry of bad studies have suggested they may exacerbate anorexia. The concern is reasonable but taken together the papers — like those used to justify calorie counts in the first place — sum up to intellectual fashion, not knowledge. That academics and policy makers are content with poor evidence doesn’t mean such evidence is good enough, but that those in charge of society aren’t serious about their decision making.
What matters in medicine is not theory but practice, and to be known, practice must be measured. The steady reduction of people’s freedom to smoke tobacco seems to me to be justified not because it is morally admirable but because it prevents vast numbers of premature deaths. Surgical masks to prevent covid were not objectionable because they were intrusions, but because the evidence showed they didn’t work. I loathe the recent proliferation of 20mph speed limits, but if they can be shown to stop a large number of road deaths then I am willing to swallow my dislike.
We are in the midst of a global expansion of public health intrusions, and in itself that is no bad thing. Lives are becoming more precious, interventions more powerful. But the existence of helpful interventions has brought about the presupposition that all interventions are helpful, and they aren’t. It is worth remembering that calorie counts on menus were not just the idiosyncratic relic of Boris Johnson’s tenure — they have been introduced across the world by parties across the political spectrum. They are a manifestation of two global trends. Our urge to fight obesity is one, and is laudable. But the increasing role of government, which encourages doctors like myself to intrude ever more into the daily business of our lives, is the other, and it is more ambiguous.
Demanding that public health policies are held to a high standard should be common ground for libertarians and nanny-staters alike. They matter too much to be picked only for the apparent virtuousness of their intentions. Policy choices, like pills, deserve to be taken seriously. Both have to be tested to see what they do, not chosen on the basis of hope and faith and guesswork.
Macaulay said puritans hated bear-baiting not because it gave pain to the bear but pleasure to the spectators. Lacking robust evidence to justify their existence, calorie counts on menus are thoughtless puritanism, masquerading as science. They make the overweight feel bad, and allow politicians and public health officials to feel good. In themselves they probably matter little. As a demonstration of our lack of seriousness when it comes to evaluating policy, they matter a great deal.
