Evolutionarily speaking, we are as a species hardwired to analyze risk based off of information that's directly in front of us--immediately accessible to our five senses. We're designed not to trust food that smells funny, can instantly calculate how far away we should stay from large cats capable of having us for a snack, and do a host of other things that were very useful for us to eke out a living in Olduvai Gorge.
But we live in the 21st century, and nowadays our ability to perceive and estimate risk is hampered by the fact that many of today's risks are abstract, and require a resonably sophisticated understanding of statistics. Take, for example, a recent discussion found in the Paper Of Record about income distribution in the United States. True--it's not really a round table about "risk" per se, unless you consider radically unequal wealth distribution to be a risk to democracy, as Supreme Court Justice Louis Brandeis did when he said that "we can have democracy in this country, or we can have great wealth concentrated in the hands of a few, but we can't have both." Still, the NY Times roundtable was remarkable in that all of the contributors, whether approaching the issue from a left or right viewpoint, agreed that most Americans had vastly underestimated how much wealth is held by relatively few. In particular, a study by Michael Norton and Dan Ariely found that not only do Americans think that wealth distribution to be significantly more equitable than it actually is, but that they would prefer it to be even more equitable than what they (wrongly) perceive.
If this isn't a classic example of what George W. Bush would call "misunderestimation" then it's not clear what is, and moreover, it highlights the difficulties people have in making accurate estimates about things like the distribution of wealth in a hugely complex society: the information simply cannot be found by opening your eyes and looking around. In medicine, we see this all the time: people are often terrified of exotic diseases that pose little threat to them, while being utterly blithe to the daily assaults on their bodies--frequently self-inflicted--that are much more likely to send them six feet under. To wit: drinking, smoking, eating poorly and not exercising.
The recent events at the Fukushima Daiichi nuclear power plant have been a case study in this process of risk assessment, and not altogether surprisingly, we haven't done well collectively in harmonizing our level of panic to the actual threat that the reactors pose. Despite a good number of depressing news stories, some cataloging evasive action by non-Japanese governments, it is far from clear how huge an impact the nuclear accident is going to have. While it is already comparable to the Three Mile Island accident in 1979, and is not (yet) as catastrophic as the Chernobyl accident of 1986, the question still remains: just how dangerous is it? Though the story is far from over there with events taking dramatic swings in short periods, the short answer is something like dangerous, but not as dangerous as you think. Not nearly as dangerous as you think.
That point is neatly illustrated, both in sound and visual format, in this news story from Adam Ragusea of WBUR (Boston), and is described as the "Dread-to-Risk ratio" by Andrew Revkin of the Dot Earth blog at NYT. Both pieces have the same useful graphic to give you some sense of the relative levels of radiation that we're talking about. Live within 50 miles of a nuclear power plant for one full year? That will give you about 0.1 microSieverts (uSv) of radiation. (What a Sievert is, is a longer discussion, but we'll just shorthand it here and say that it's some relative value of radiation, and that the higher the number, the more dangerous it gets.) One flight from New York to LA buys you about 400 times that amount (40 uSv). That's not even the round trip! A standard chest x-ray, meanwhile, is worth about 20 uSv. But a mammogram is a whopper, clocking in at 3 milliSieverts--thus about 150 "standard" x-rays and just under forty roundtrip flights from NY to LA. A CT scan can be worth almost twice the amount of the mammogram (5.8 mSv).
(In case I've lost some people on this micro/milli distinction, you need 1000 "micros" to make 1 "milli." I'm going to flip back and forth but will point it out when I do.)
So how do these numbers stack up to the nuclear disasters? If you lived within 10 miles of the Three Mile Island plant during the accident and didn't make a run for it, the total dose of radiation you received was 80 microSieverts--far less than one mammogram. By contrast, one area near the Fukushima plant recorded a total dose in one day of 3.6 milliSieverts: less than a CT but more than a mammogram, though of course we're only talking about one day's worth of radiation. With Chernobyl, the radiation levels fluctuated wildly both in time and place so making a general statement about the radiation is essentially impossible, but had you been moved by some weird spirit to take a stroll on the grounds just last year, about 25 years after the accident, you would have gotten two mammograms' worth of radiation for your troubles: 6 milliSieverts. I won't reproduce the pic here out of respect for copyright but highly recommend it to anyone with the time; Ragusea of WBUR translates this into a tone equivalent, and the radiation from TMI is a blip, while the sound for "mammogram" is substantially longer.
I draw two conclusions from all of this. First, while the troubles at Fukushima are by no means trivial, and for that matter aren't yet finished, I think it's a bit premature to write the obituary for nuclear power. In terms of accidents, it's not nearly as dangerous as most people suppose. The problem with nukes, in the TMI and Chernobyl age as well as today, is what to do with the radioactive waste generated by the plants rather than the risks they pose viz. accidents. Second is that we should try to minimize mammograms! Do only the amount that will help save lives, and not one more after that. This was the logic behind the recently revised US Preventative Services Task Force, which recommended no mammograms to women under 50, and biannual ones to those over. Despite this entirely sensible approach--based on good research with a careful eye toward the risk/benefit ratio of the radiation, it should be noted--there were howls of indignation from people purportedly speaking for women, accusing the very bureaucrats who issued the new recs to be female-hostile, or something like that.
(NB--the first draft of this version, which snuck out prematurely, posted some incorrect calculations with respect to x-rays, mammograms, and NY-LA flights. The corrected version is now present.)