This is a good article, and it's unfortuntate that only two people (including me) have clapped for it.
However, arguing against LNT specifically is, I think, a mistake. I'm not arguing that LNT is true, and I'm not even arguing that LNT isn't clearly false (although yes, it isn't totally clear to me that LNT is false). But I don't think it's the right thing to argue about.
Let's consider the humble hot dog for a moment. The WHO classifies hot dogs as a carcinogen.
A lot of people will knee-jerk react that "oh, the WHO must not know what it's talking about. Ridiculous! Hot dogs? Carcinogenic? Please."
People react this way not because they have a body of medical evidence that hot dogs don't cause cancer. It's because they like hot dogs. But that's not a rational reaction, is it?
A more rational reaction is this: you see that it's a very mild carcinogen - so that eating a lot of hot dogs gives you a very tiny chance of dying - and you say "okay, but hot dogs taste really good, so I can live with a tiny chance they will kill me, but I will try not to eat too many hot dogs."
Let's also consider airplanes. Is it possible to die in a plane crash? Yes. Would you be willing to pay twice as much for plane tickets, in order to reduce the risk by half? No, because the risk is too small to worry about. (If regulators had forced ticket prices way up in exchange for excessive safety measures, it may have decreased per-passenger airplane safety via reduced demand, but that's a different topic.)
The risks of airplanes and hotdogs might well be linear-no-threshold: your risk increases in proportion to the number of flights and number of hot dogs.
But if the risk is small enough, then it's okay. You are geting something you want in exchange for that risk of death. People have to live their lives; minimizing the risk "as low as reasonably achievable" means not leaving your house and not eating the food you like.
In the case of radiation, it matters what the coefficient is — how much danger does radiation pose? This question is almost always left unanswered, including by this article. After Fukushima they eventually (after causing tremendous disruption to relocated people's lives) said that people could move back if radiation levels dropped below 20 mSv per year.
Well, is 20 mSv the "right" cutoff? (Remember, this is a coerced limit: people were not "recommended against" going home if there was 21 mSv per year; rather, going home was outright prohibited.)
Our judgement here depends not so much on whether LNT is true or false, but simply what our risk of cancer would be. Are we talking about a 1-in-100 chance of cancer? 1 in 1000? 1 in 10,000? 1 in 100,000? If the chance is one in a hundred thousand, that should be a hundred times less scary than 1 in 1000, which in turn should be ten times less scary than 1 in 100. And yet, virtually no one — pro-nuclear and anti-nuclear alike — tries to report what that risk is! Instead we have anti-nukes using scary language and talking about non-dose units like Bq that sound scary because they are big numbers but don't tell us anything about absorbed dose, let alone risk. Meanwhile, pro-nukes don't have the information so they don't report it.
Granted, it's hard to estimate the risk because we've worked so hard to minimize radiation exposure, across the entire world, that very few people have ever recieved 20 mSv per year! 20 mSv, by the way, was approximately the amount of radiation I received in a recent cardiac function test that used some kind of gamma-ray tracer, so you'd think we could do a study of medical tests as a way of estimating the risk. It's not that simple, though, first because that 20 mSv dose is received all at once (so the effect of that is not guaranteed to be the same as 20 mSv spread out over a year), but more importantly, because doctors order tests when they think something is wrong with you. So if you compare people who had tests with people who didn't have tests, the people who didn't have tests are probably healthier, but not because the tests were harmful. I'm not sure how researchers deal with this problem. I certainly have the impression that enough research has been done to estimate the risks of 20 mSv or 100 mSv or whatever amount of radiation, but that information is almost never disseminated in a meaningful form . So if you could find that information, I for one would love to see it, and it would certainly be more important than whether LNT is true or false.
If the "linear threshold" model were approximately true, and if the threshold were 25 mSV per year, then 20 mSV would leave you with no risk at all. But that's only relevant for doses under the threshold, and (1) I've never heard anyone tell me what exactly the threshold might be and (2) honestly I don't think that the threshold, if there is one, is as high as 20 or 25 mSV/yr. If you get a dose larger than the threshold, then you suddenly care very much about the coefficient, like, do I get a 1-in-1000 risk of cancer for each additional mSv, or is it more like one in a million? In summary, just knowing whether there is a threshold or not — but not knowing what the threshold is or the slope of the line — is not very useful.
The other issue I'd like to bring up is that comparative badness matters. That is, given some large amount of energy, lets say one TWh of energy, how many deaths are caused by using one energy source instead of another?
Everybody knows that Chernobyl caused, or will cause, up to 4,000 deaths (or 9,000, or 16,000 depending on which study you look at) but the media typically does not compare this number with the number for fossil fuels, which is hundreds of times more deaths. So this chart belongs in any article promoting nuclear power IMO: https://ourworldindata.org/grapher/death-rates-from-energy-production-per-twh
If we take one nuclear plant offline, where do we get electricity from instead? The answer is that more fossil fuels will be burned immediately (causing more deaths). But hey, one could argue today that renewables will eventually replace that demand. However, 50 years ago the answer was more stark: preventing a nuclear plant from being built would mean that a coal, oil, or gas plant would be built instead, and that 50+ years of fossil fuels would be burned instead. And yet, activists were out in force 50 years ago to oppose nuclear plants but not coal plants. And so, today, global warming is a big problem and millions more people died early than otherwise would have. And the arguments they were using 50 years ago are the same arguments they use today, which makes me suspicious. (The only difference is that today anti-nukes use a new argument that nuclear is too expensive. But why is it so expensive? Because of the new regulations that they themselves demanded, regulations that they do not recognize today as increasing safety.)
Even today Germany's plan was to phase out nuclear this year, several years before phasing out coal. And after Russia invaded Ukraine, Germany was quick to create a plan to burn more coal to reduce reliance on Russian gas, but slow to delay the closing of nuclear plants. It is clear, then, that anti-nuclear forces 50 years ago, as well as today, preferred coal over nuclear power, even though the environmental and health effects of coal are hugely, obviously, immensely worse than nuclear power. What accounts for this? My theory has to do with nuclear weapons: they have mistaken beliefs such as (1) nuclear plants can blow up like nuclear weapons, (2) all reactors are alike, implying that any western reactors could blow up like Chernobyl, (3) any power reactor can be used to help make nuclear weapons, (4) nuclear waste will inevitably be extremely dangerous for over 10,000 years. All of this is simply untrue.
So, respected organizations consider both radiation and hot dogs to be carcinogenic. But there's a massive, 60-year-old, very successful political movement to create enough fear to make nuclear power die, and there is no similar movement for hot dogs. Ultimately, that's why we can simply buy a hot dog or a plane ticket — both at reasonable prices — and not worry so much about it.
