For many years, public health people have thought that if only they provided figures on the cost of poor diets, policymakers would respond. Reviewing national figures for a summer school this week, and noting the snail-like pace of change in policy, food supply and consumer behaviour, I now doubt this logic. ‘Since when was policy based solely on evidence?’ I hear you ask.

Consider the data. Food-related diseases account for an annual 125,000 premature deaths before 75 years, two-thirds of all preventable deaths - 30,000 of these could be prevented if people ate currently recommended diets.

Only 24% of men and 29% of women consume their five-a-day. Only 16% of boys and 20% of girls. Fresh and processed fruit purchases have dropped since 2008. Only 36% of adults do moderate-intensity physical activity at least once a week.

“The food system makes some changes, but only tiny ones”

The cost of heart disease to the NHS alone is at least £6bn a year. Hidden costs to families and workplaces add much more. Food poisoning costs an estimated £1.5bn in total, both to the NHS, and work and families.

There’s some good news. Heart attack deaths have dropped - and not just because of quicker ambulances and better healthcare. About half the drop in deaths is due to people changing lifestyles. Prevention works.

Such data can be developed for many aspects of the food chain. Asked to turn everything into costs and direct impact, scientists and academics have painstakingly done so. Biodiversity loss, hunger, water use - you name it. Food accounts for about 30% of climate change impacts by average European consumers. So are we cutting meat and dairy?

True, the food system makes some changes, but only tiny, incremental ones. The pace is too slow and the scale too narrow.

Is it only shock that genuinely gets systemic change? In business, it seems so. Witness Tesco’s internal traumas from horsemeat this year, or consider how farmers learned they’d been feeding recycled animals to cows, creating BSE. By the way, there’s good news here. At its 1992 peak, there were 36,000 recorded cases of BSE, in 2012 there were two and so far this year just one.

So how many deaths and costs does it take to change?

Tim Lang is professor of food policy at City University London