Fukushima babies and how numbers can lie


Xeni brought this to my attention yesterday. Over at Scientific American, Michael Moyer takes a critical look at an Al Jazeera story about a recent study purporting to show that infant deaths on the American West Coast increased by 35% as a result of fallout from the Fukushima Daiichi nuclear power plant meltdown.

At first glance, the story looks credible. And scary. The information comes from a physician, Janette Sherman MD, and epidemiologist Joseph Mangano, who got their data from the Centers for Disease Control and Prevention's Morbidity and Mortality Weekly Reports—a newsletter that frequently helps public health officials spot trends in death and illness.

Look closer, though, and the credibility vanishes. For one thing, this isn't a formal scientific study and Sherman and Mangano didn't publish their findings in a peer-reviewed journal, or even on a science blog. Instead, all of this comes from an essay the two wrote for Counter Punch, a political newsletter. And when Sci Am's Moyer holds that essay up to the standards of scientific research, its scary conclusions fall apart.

Sherman and Mangano tally up all the deaths of babies under one year old in eight West Coast cities: Seattle; Portland, Ore.; Boise, Idaho; and the California cities of San Francisco, Sacramento, San Jose, Santa Cruz and Berkeley. They then compare the average number of deaths per week for the four weeks preceding the disaster with the 10 weeks following. The jump--from 9.25 to 12.5 deaths per week--is "statistically significant," the authors report.

Let's first consider the data that the authors left out of their analysis. It's hard to understand why the authors stopped at these eight cities. Why include Boise but not Tacoma? Or Spokane? Both have about the same size population as Boise, they're closer to Japan, and the CDC includes data from Tacoma and Spokane in the weekly reports.

More important, why did the authors choose to use only the four weeks preceding the Fukushima disaster? Here is where we begin to pick up a whiff of data fixing. ... While it certainly is true that there were fewer deaths in the four weeks leading up to Fukushima than there have been in the 10 weeks following, the entire year has seen no overall trend. When I plotted a best-fit line to the data, Excel calculated a very slight decrease in the infant mortality rate. Only by explicitly excluding data from January and February were Sherman and Mangano able to froth up their specious statistical scaremongering.

You can see that data all plotted out nicely by Moyer over at Sci Am. How did these numbers get so heinously distorted? It's hard to say. But there should be some important lessons here. In particular, this is a good reminder that human beings do not always behave the way some economists think we do. We're not totally rational creatures. And profit motive is not the only factor driving our choices.

When you think about what information be skeptical of, that decision can't begin and end with "corporate interests." Yes, those sources often give you bad information. But bad information comes from other places, too. The Fukushima accident was worse than TEPCO wanted people to believe when it first happened. Radiation isn't healthy for you, and there are people (plant workers, emergency crews, people who lived nearby) who will be dealing with the effects of Fukushima for years to come. But the fact that all of that is true does not mean that we should uncritically accept it when somebody says that radiation from Fukushima is killing babies in the United States. Just because the corporate interests are in the wrong doesn't mean that every claim against them is true.

Reality leaves us in a confusing place. As Slacktivist, one of my favorite bloggers, likes to say, "It's more complicated than that." With "it," in this case, meaning "everything." The best solution is to always ask questions, even when the people we're questioning look like the good guys.

Image: Baby toes, a Creative Commons Attribution (2.0) image from 40765798@N00's photostream


  1. Counterpunch itself investigated the claim and found:

    “Simply by moving the boundary line northward from Santa Cruz Sprey found that the four northernmost Pacific Northwest cities in the CDC sample – Portland, Tacoma, Seattle and Spokane – show remarkably significant results – a larger infant mortality increase than the original Sherman-Mangano results.

    During the ten weeks before March 11 those four cities suffered 55 deaths among infants less than one year old. In the ten weeks after Fukushima 78 infants died – a 42 per cent increase and one that is statistically significant. To confirm once again that these results were not due to seasonality Sprey compared these infant deaths in the ten weeks after Fukushima to the deaths in the equivalent ten weeks a year earlier. The results were almost identical with the ten weeks before Fukushima in 2011. Within the equivalent ten weeks of 2010 53 infants died in these four cities.

    The post-Fukushima deaths are 47 per cent higher than they were in the same period a year before – once again statistically significant. If you add Boisie, Idaho to the four city sample the results remain almost unchanged.

    Looking a little more closely at the time trend of the infant deaths after Fukushima, Sprey found that the most dramatic increases in deaths were in the two weeks right after the March 11 disaster. Those two weeks saw a near tripling of weekly deaths, followed by a period of somewhat elevated weekly deaths lasting for about five weeks – roughly 25 per over the pre-March 11 rate, then settling down close to the average pre-Fukushima death rate for the last three weeks of the ten week period post-disaster. These results are necessarily approximate because the weekly sample of deaths is too small for precise statistical conclusions. ”

    Bottom of page:

    This is far from a settled issue.

  2. PS

    The Scientific American story introduces the same error as it claims to disprove. He arbitrarily chose to start with January. Why not December? Why not November, etc?

    The only thing we learn from this is that much more investigation is needed. It’s no secret that radiation is harmful to fetuses.

  3. Yay, statistics debate! Shouldn’t the four week period after the Fukushima disaster been compared to the same period one year before? Studying only the four weeks prior to the disaster misses the potential for seasonal variation in infant deaths (although I’m not sure if there is seasonal variation, I’m just throwing this out there).

    1. OMG! Statistics debate!! OMG!
      I’m so excited…OMG!…geez, I’m having trouble breathing here…I gotta go lay down for a moment…whew!…
      A statistics debate!! Wow!

      1. Bah, I meant X and Y, not a and b, in my comment.

        Stinking variables, looking so confusingly similar and all.

  4. Deaths per week is meaningless, since births aren’t evenly distributed over the year. Why didn’t they use deaths per birth, Negligence or malice?

  5. I’m somebody who has worked with antinuclear activists and is generally very critical of nuclear power, and I think that articles like what ran in Al Jazeera are very unhelpful. What I’m concerned about has to do with what radiation levels workers at Fukushima are exposed to, and the exposure for children in cities not in the evacuation zones (which are about 10x background in some places.) The idea that you could have a sudden massive increase in immediate health effects thousands of miles away just lowers the level of debate and assures that things stay at the level of paranoia vs. denial.

  6. 90% of journalist don’t understand statistics.
    83% of statistics are made up on the spot.

    1. “90% of journalist don’t understand statistics.
      83% of statistics are made up on the spot.”

      271% of commenters on statistics stories don’t understand stats

  7. But to point out a counter issue, what are these deaths?

    Does the CDC have statistics on type of death? Accident, SID, ingestion of toxic substance, choking, drowning, ect..? If so wouldn’t it make more sense to eliminate all the things that aren’t radiation related?

    If anything wouldn’t it be more important to note the number of cases of cancer in children since this all started?

    On a personal note, I have a really hard time believing radiation from a reactor in Japan caused a higher number of children to die that far away. Did it cause damage, maybe, but not on that kind of scale or time table.

    1. the baby deaths were too soon after the fallout, to be from cancer, since cancer usually takes many months or years to develop, not just a couple weeks. the comment below about seattle hospitals finding super high rate of thyroid pathology in newborns might give more of a hint as to some likely effects.

      (getting “comment submission error, so attempted to post more than once).

  8. Not only is Spokane closer to Japan than Boise but it is also in the direct jetstream path from Japan. Which is why some of the rice paper balloon bombs landed nearby during WWII. Which would lend one to believe that would be in the fallout path.

  9. Honestly, I’m really much more interested in the total news blackout surrounding the three US nuclear plants that are underwater right now.

    What’s being done about the on-site waste storage situation, to prevent the waste from traveling down the Mississippi into densely populated areas? What’s the shutdown status; are they cold, as has been claimed, or hot, as might be indicated by reports from eyewitnesses who illegally flew into the no-fly zones the corporations have established to keep out news choppers? Are the dikes holding with water 3 inches from their tops?

    Can we get some information on that situation?

    Yeah, I didn’t think so.

    I guess we’ll just have to talk about how safe we are from Japan’s mistakes, instead of looking at the nuclear plants in flood zones and on fault lines in the USA.

  10. Thanks, Maggie, for reminding us that reversed stupidity is not intelligence. If it were, we could just hook the dumbest person in the world up to a NOT gate.

    @#3: I did
    @#4: The point of the SA article, I think, was to make clear that the original claim was based on arbitrary and unjustified assumptions. AFAIK, it did not claim to eliminate these and perform a flawless analysis.

  11. “Looking a little more closely at the time trend of the infant deaths after Fukushima, Sprey found that the most dramatic increases in deaths were in the two weeks right after the March 11 disaster.”

    The fact that radiation didn’t show up in Pacific Northwest for over two weeks after the disaster actually tends to prove the radiation reduced baby deaths.

  12. Whenever something like this comes up I always remind myself that a book titled ‘How To Lie With Statistic’ was an optional text in my first college stat class. Then I repeat the mantra ‘correlation is not causation’ several hundred times…

  13. “The only part of an argument that really matters is what we think of the people arguing. X claims a, Y claims b. They make arguments to support their claims with any number of points. But when their listeners remember the discussion, what matters is simply that X believes a and Y believes b. People then form their judgment on what they think of X and Y.”
    — Kim Stanley Robinson (Red Mars)

    1. Make a and b identical twins, identically dressed; or, have a and b present their arguments through the medium of a herald, who voices both arguments.

      That should allow people to focus on the matter at hand, and ignore the presenter(s) – as the presenter is the same for both sides.

  14. BB has continually sought to downplay the dangers of Fukushima.
    Was this study a game of statistics? I can’t say, and perhaps it was.
    What I can say is that the same kinds of statistical games are played with global warming, and BB never questions the methodology of those.

    Furthermore there is perhaps good reason to have chosen the areas chosen. Distance is NOT one of them, but then distance is NEVER a good predictor of fallout after a nuclear accident… just look at the irridation maps from Chernobyl and see what I mean.

    The Pacific NW, Pennsylvania and Delaware had unusually high levels of radio-isotopes in their rainwater, as measured by the EPA, shortly after the explosions. I presume that these cities were chosen for their proximity to these known high readings. Of course, they could have just split up the nation however they wished until they found locations suiting their needs- the global warming “scientists” do that all the time. At least there seems no evidence that these researches “added value” to the data, which the AGW crowd also does, and again no one in the mainstream media questions that either.

  15. If you read the Al Jazeera article, it’s chock full of scaremongering. I dug around and looked at the original Counter Punch article, which of course had no more details.

    I wasn’t about to spend the time debunking it, but I’m glad to see someone (with rather more credentials than me) doing so.

  16. I know someone who works with newborn babies at a hospital here in Seattle. I’m told that while thyroid function is a routine check that is given to all these kids, a bad test result is almost unheard of.

    But by some strange, unknowable coincidence, in the last few weeks the tests have been coming back bad in about 1/3 of cases.

    I find it quite credible to think that babies here in the states are suffering for what’s happened on the other side the ocean. What would be unthinkable would be for health officials to sound the alarm and confirm that there is a risk- it’s not like there’s anything we could *do* about it, after all!

    1. Once those numbers are actually published, we’ll know the extent of the problem, but until then, it’s just speculation, and rumour.

      Here’s some news about some actual peer reviewed science, as to birth defects in areas near mountain-top removal coal operations in the Appalachians:


      Something to reflect on, until the Seattle “radiation birth defect” observations are presented in a more complete and reliable form, perhaps.

  17. Yes, Al Jazeera and some political blog called Counter Punch, that’s where I get my facts and science from!

    Thanks for digging and debunking -even though they discredited themselves from the very start by how (poor stats, cherry picking data, poor study design, etc.) and where (not in any sort of peer-reviewed or even other respected scientific publication) this information was presented.

  18. I’m doing some research to post a bit on that report in the next couple weeks.

    Right now, there is very, very, very little known definitively about what is going on with infant mortality rates on the U.S. West Coast. And what we do know for sure cannot (at this point) be tied to radioactive fallout.

    I’m all for calling out spikes in infant mortality when they happen. But I also happen to think it’s pretty unnecessary (and, in fact, somewhat discrediting of situations where the evidence is strong) to take largely undocumented phenomenon of unknown origin and call it a result of radioactive fallout.

  19. Bravo. Because the statistics about the side-effects of careless operation of nuclear plants don’t need help from bad science.

    This is in fact just the sort of thing that might be deliberately planted so that it can be shot down later. Thus (as in the case of global warming) allowing interested powers to gain consent for perfidy.

    Of course, getting good numbers and good science about nuclear side-effects can be very, very hard indeed. It doesn’t help that the plethora of exposure measures (curies, sieverts, rads, rems, becquerels) make it all difficult to grasp. It doesn’t help that few media outlets with a large audience dare to risk provoking the wrath of such an influential lobby. It doesn’t help that professionals can shit-can their careers by choosing to get anywhere near the facts. And it doesn’t help that decades of deceptions and coverups (just like TEPCO) make it easier to turn away and “let the experts worry about it”.

    It’s easy to see why people who really care, short on time and cash, sometimes go overboard. Much easier to forgive than the cynical, insidious, perpetually funded manipulations of the people who just don’t give a crap.

  20. Ok, I havent read all the comments, i havent read the book ..

    The 10 or 4 week problem isnt the biggest problem to me. Thing is that evolution of child mortality could be linked with many seasionial factors that might have not been considered here (temperature, humidity, diseases … etc).

    In order to set this right, they should have study evolution of child mortality in the same 14 weeks period in 5 years or more.

    Or maybe they have …

  21. I don’t care about the statistics you posted. It’s a tough thing to analyze just 3 months after the disaster. I need to say this: To even suggest the possibility that there may NOT be any increase in infant mortality overseas due to Fukushima is complete nonsense. A small radiation dose spread out among the whole population WILL increase the statistical likelihood of birth defects (which can lead to complications and death), cancer, and other types of tumours and mutations. I was born prematurely, in the months following Chernobyl, with a handful of rare birth defects which aren’t typically co-morbid. I’m on the same continent as you, but I apparently wasn’t immune to the 325 millibequerels in Ontario rainwater from may to the end of june in 1986. I search the internet for the causes of my conditions dextrocardia and fibrous dysplasia and all I find is this convenient term the scientific community uses, “genetic damage”. It seems they don’t want to admit that radiation causes the genetic damage that causes these conditions, but it seems obvious to me. I was one of those babies where the doctor said, upon my birth, that I wouldn’t live for very long. I weighed 3 pounds. I’m lucky. I was almost one of the unknown infant victims of a nuclear disaster. Until April 86, mine was a healthy pregnancy, though two months later I came out as a 7-month term C-section in early July riddled with health problems. My whole life has been nothing but health problem after health problem, but I’m alive. Unlike all of the babies that Chernobyl killed, and that Fukushima will kill, I am actually able to sit here and tell you to leave the science to the scientists, please and thank you, and stop spreading the seed of misinformation. There’s more important things to be dealing with regarding this disaster, such as the HEALTH CONSEQUENCES. Rather than debating whether or not there will be health consequences. (news flash: there will be health consequences)

  22. There are statistics programs that will do this kind of data-mining for you. It’s considered to be a very helpful tool, something that will find trends in your data that you might miss otherwise. These programs do not require a degree in statistics to run, and at no point do they flash warnings about how certain kinds of cherry-picking are adding a bias to the model. If you go into your analysis with the idea that you KNOW there is going to be an effect, and don’t find one right away, this is the next obvious step to take: just “do more work”, and your cleverness and tenacity will lead to important findings. And how could that be a bad thing?

    This kind of analysis is done often in Washington by public policy / lobbyist firms. And many of them don’t consider their methods to be at all suspect, and they deride the findings of other groups that use the same data sources to come to opposite conclusions.

    Others, of course, are completely aware of how they are twisting the data when they do this kind of data-mining.

  23. No matter what you think of nuclear power; the Sherman-Mangano data are cherry-picked, as concluded in Scientific American.
    Here is my version on the same thing:

    And there is something seriously wrong with the CounterPunch reanalysis:

    There may be many nasty facts to discuss in relation to nuclear power and Fukushima. This is not one of them. The only thing to discuss here is: why would an M.D. like Janette Sherman intentionally cherry pick data in this way? And if it is accidental, it only shows that she and Joe Mangano are not serious in how they perform their research. Thereby they put all their earlier works in serious doubt.

  24. correlation does not automatically imply causation

    what if there really is an increase but the increase was limited to infants born after the disaster? If they were in the mother’s womb during the disaster then the mothers could have stress hormones in their blood from watching CNN and the stress hormones are the cause?

    There are many other possibilities

  25. Pregnant mothers in Japan were advised to evacuate early on. Many Japanese citizens and Amerian citizens who had been in japan at the time of the disaster CAME to the US West Coast, and delivered here. With this fact in mind, the reported data is seriously skewed.

  26. One of the reasons we find so much dissension and debate, is that people tend to get too deeply analytical. The forests is lost for the trees. We could go back to the Story about Solomon where he asked the two women who claimed the baby if he could cut it in half: the true mother said no, she can have it.
    Here, we simply ask anyone who doubts the statistics in play, if they would take their children any closer to Fukushima if we gave them free rent and a 100,000 a years stipend. Of course, there is a 30 kilometer circle around the place where no one is allowed. So it’s really just a matter of How large is the circle?
    I think it will get larger as long as the plant is leaking radioactivity–and don’t forget the stuff is cumulative. Even Iodine, which has a half life of 8 days, if i remember right, will build up under constant replenishment. So what this small tree of info tells us is there is a problem coming, for sure. Maybe it’s now. Maybe there was a mosqutio up there in Boise with anew virus—but i’ll place my bet–be it my childrens’ lives–on looking to get out of Dodge. And writing my stupid GOV. in a fury to shut down that loan for more nukes in Texas (to be run by TEPCO, BTW.)
    Germany, Switzerland, and now Japan have all sworn off these plants and will begin to dismantle them in favor of wind.
    How many kids will have to die for this centralized power greed?
    I met a guy who asked me, what do you care? You got not so much longer to live anyway!
    i said that kind of thinking is what got us to here.
    Evacualting the children from the West Coast of the US ought to finally send a permanent indelible message to these pigs.

    1. Anon #41:

      On the contrary, I think the reason there is so much ‘dissension and debate’ is that many simply do not know what they are talking about, and have little or no evidence of the truth of their stated beliefs.

      Here’s some facts:




      I think you’d be surprised at how many would immediately return to live in the presently-evacuated zones, if the Japanese Government would let them do so.

      I add, that I consider it wisely prudent of that Government not to allow them to so to do, for the present.

Comments are closed.