Dmitry Kosyrev emailed me today wondering where all the millions of dead smokers come from. He asked:
I want to know how do they count all these dead people in millions.
And I replied:
There are no ‘millions of dead people’. There are no morgues full of dead smokers. In fact, there are no bodies being counted at all.
They are all entirely imaginary dead people, conjured up using mathematics.
Step 1: You get a ‘relative risk’ from some study of children dying from, say, secondhand smoke. A relative risk of 1 is no risk at all, or same risk as everything else. A relative risk less than 1 is safer, and a relative risk greater than 1 is less safe. Let’s suppose that the relative risk is 1.13. That’s a 13% greater risk.
Step 2: 13% doesn’t seem much of an enhanced risk, so you now need the dead bodies. In your country in question, let’s say there are 13.6 million children. If these children have a 13% increased risk of dying due from secondhand smoke, then (here comes the mathematics!) that’s 0.13 times 13600000 = 1768000 dead children. Nearly 1.8 million dead children!!! How shocking!
Step 3: Now that you’ve got 1.8 million dead children (nearly TWO million!) you compare the number with real deaths from road accidents, train crashes, etc. These numbers are of actual dead bodies lying on slabs in morgues. Let’s say that there are 51,200 road accident victims, 2300 air crash victims, 786 drownings. So then you write an article in a newspaper that says that More Children Die Each Year From Secondhand Smoke Than From All Other Causes Put Together. The newspapers won’t check your numbers. Reporters don’t know how to multiply.
Step 4 (optional): You may worry that you’ve got too many dead children. It may seem slightly implausible that nearly 2 million children have died from secondhand smoke. No problem. Just go back to Step 1 and choose a smaller relative risk. Say 1.003. Now 0.003 x 13600000 = 40800. Over 40,000 dead children!!! That’s almost as many people as are killed in road accidents. But a much more plausible-sounding number.
Do you get the idea? If pushed, they’d probably be able to produce studies which show a small relative risk from secondhand smoke, and there probably actually are 13 million children. They can probably back up all their numbers. The deceit is essentially one of passing off projected numbers as being actual measured numbers, or at least leaving people with the impression that they’ve been measured, and that there are millions of dead people lying in morgues with labels tied to their big toes saying “Killed by secondhand smoke”. There aren’t any. In fact there aren’t any with labels saying “Killed by smoking” either.
It’s probably slightly more complicated than this, but this is my understanding of where all the ‘millions of dead people’ come from. If you’re in doubt, ask Michael McFadden. I’m sure he’ll provide a similar explanation…
In retrospect, I could have explained in a ‘Step 0’ how the relative risks are found.
Step 0: You want to find out how great is the risk to the lives of children from living in a smoking household. First you find a control group of 20 non-smoking households, and find out how many dead children they’ve got buried in their back yards. Let’s say the 20 non-smoking households have got 30 dead children. You then find 20 smoking households, and you find how many dead children they’ve got. Let’s say that they’ve got 31 dead children. So the relative risk to children from living in a smoking household is 31/30 or 1.033, assuming that you’ve controlled for other causes of death like hand guns, kitchen knives, and rolling pins, by ensuring that all households have about equal numbers of these. And there’s your relative risk: 1.033
It’s a bit more complicated than this of course, but essentially this is how the dead are invented. And of course if you can reduce the number of these imaginary dead with smoking bans, then you’re “saving lives”, of course.