The UN did as only the UN can do, putting out a supposedly authoritative scientific report on a “staggering” rise in climate disasters over the last 20 years that was seized upon with enthusiasm by the usual suspects. Regrettably the report itself turned out to be a climate disaster when it was pointed out that according to the data in the report disasters have been declining globally over the past 20 years. Calls for the report’s retraction went unheeded, naturally, the fact that the conclusions were the exact opposite of the evidence being considered insufficient grounds for the UN to admit error.

If you’re wondering how they could be so stupid, it’s tempting to say lots and lots of practice. And to point to the undoubted fact that climate alarmists have become so accustomed to running the table in the media and politically by dismissing their opponents as moral and mental defectives that they’ve become lazy. But the technical answer is that they compared 2000-2019 with an earlier period, 1980-1999, in which data was known to be collected far less rigorously, so the number of disasters in the earlier period was understated.

As Paul Homewood observed, after citing readily available and widely known details of the process, “Put simply, many more disasters are recorded nowadays because of better reporting systems. But this does not mean more are actually occurring.”

If you think the UN was cherry-picking, you’re right. If you think it was shamelessly cherry-picking, you’re completely right. As Homewood shows, the improved (and better-funded by USAID) data collection system kicked in right at the end of… well, you saw it coming, right? The 1990s. Indeed, “there is a sharp jump in 1999/2000”. Exactly the break between the two data sets the UN compared. It’s not a subtle error; indeed it is hard to convince oneself it is innocent though as so often, people very sure of what they’re going to find often manage to find it in a process of unconscious self-deception rather than the deliberate deceit of others.

Homewood’s point is underlined by one category in which we do seem to have fairly reliable data, namely deaths from natural disasters or, as the UN prefers to call them, “climate-related disasters”. Deaths are more likely to be noticed and recorded than other less serious results even of very serious events. And, he says, “Note that despite the claimed increase in disasters, the death toll has nearly halved”, from 995,300 deaths due to “3,656 climate-related events” in the earlier period to 510,837 deaths due to a reported “6,681 climate-related disasters” in the latter.

It’s revealing that the report itself refers to “events” in the earlier period and “disasters” in the latter. Because it’s not science, it’s propaganda in the worst sense of that word.