It began by trying to discover whether or not a widespread belief was true. In doing so, I tripped across something even bigger: an index of the world’s indifference. I already knew that by burning fossil fuels, gorging on meat and dairy, and failing to make even simple changes, the rich world imposes a massive burden of disaster, displacement and death on people whose responsibility for the climate crisis is minimal. What I’ve now stumbled into is the vast black hole of our ignorance about these impacts.

What I wanted to discover was whether it’s true that nine times as many of the world’s people die of cold than of heat. The figure is often used by people who want to delay climate action: if we do nothing, some maintain, fewer will die. Of course, they gloss over all the other impacts of climate breakdown: the storms, floods, droughts, fires, crop failures, disease and sea level rise. But is this claim, at least, correct?

The figure comes from a study using the widest available datasets to try to produce a global view. The results are, to say the least, surprising. For example, it suggests that even in the hottest parts of the world, more people die of cold than from heat. In fact, sub-Saharan Africa appears to have the world’s highest rate of deaths from cold and the world’s lowest rate of deaths from heat. The figures suggest that 58 times as many people there die of cold than of heat. While it’s true that in hot places people are less adapted to cold, can this really be so?

The paper explains that its dataset “covers 750 locations in 43 countries or territories”. But the only African country covered is South Africa. Nor are there any data from India, Pakistan, Bangladesh, Afghanistan, the Gulf states (except Kuwait), Indonesia or Melanesia. In other words, most of the world’s hottest countries are not represented. Nor are most of the places in which healthcare is weakest, either for the population as a whole (as in some African nations) or for the most vulnerable people (as in the Gulf states, where citizens might be well covered, but migrant workers scarcely at all). This is in no way the fault of the authors – it’s simply a matter of where records are available.