In the article Mangano is using monthly mortality statistics from the Japanese Ministry of Health, Labour and Welfare (MHLW), in order to investigate health effects due to the Fukushima nuclear disaster. This is mixed with the usual fearmongering, advertisement for his earlier work and for the infamous Chernobyl book that somehow was published by the New York Academy of Sciences. The central part of the article reads
Total deaths increased 4.8%, compared to the normal 1.5% annual rise. Since about 1.2 million Japanese people die each year, this computes to an excess of 57,900 deaths. The rise in deaths from accidents is given as 19,200, close to estimates of those killed directly by the earthquake and tsunami. But this still leaves an excess of 38,700 Japanese deaths, with no obvious cause.
There are a number of statements here that are worthy of closer scrutiny, not the least with respect to Mangano's earlier cherry-picking exercises after Fukushima (see this post and links therein).
Data from MHLW are available from the following link: http://www.mhlw.go.jp/english/new-info/. It is easy to get lost in this data base, especially for non-Japanese speakers, and there are several versions of the same data (preliminary data, projected estimates, final data). Due to this, and partly due to lack of time, the comparisons below may not come from exactly the same tables as the ones used by Mangano. Therefore absolute numbers may differ, though the trends and the relative differences (an important variable that Mangano often uses to push his agenda) should agree very well.
A dramatic increase?
Mangano states that the total deaths in Japan increased with 4.8% for 2011 (or the 12 months March 2011-February 2012), while the normal annual rise is 1.5%. It really sounds remarkable, but is it? Let us plot the annual change in total deaths in Japan (i.e. the number of deaths for a given year divided with the number of deaths for the year before), below is a plot with all data from the year 1900 until 2011.
Considering the effects of wars, economic and demographic changes, improvements in health care, etc., it is meaningless to discuss a normal annual rise in the number of deaths over this long time period. But the plot shows that historically the annual changes have been quite large. There are actually a few outliers larger than 10% (positive as well as negative) that are not displayed in the plot. The next plot shows the same data but only since 1990 and later:
The red dashed line shows the average increase over the time period 1990-2011, roughly 2%, so Mangano is not so far off with his 1.5% (see the appendix in the bottom on why there should be an increase in the number of deaths). But we can see that the annual change varies quite drastically, 1.5% or 2% increase is by no means a "normal" number that we should expect every year. And we see that during the last 20 years the increase in number of deaths from one year to the next has reached 4-5% six times (1990, 1995, 1999, 2005, 2010 and 2011). To my knowledge it is only during 2011 that there were nuclear meltdowns, something worth remembering for those prone on blaming radioactivity. The fact that the increase in 2010 was as high, or actually slightly higher, makes Mangano's scaremongering phrase "an excess of 38,700 Japanese deaths, with no obvious cause" sound quite silly. Anyhow, Mangano corrects himself immediately, though the "no obvious cause" phrase has given the impression that we are dealing with large numbers:
The reports provide mortality numbers for 12 common causes, making up about 80% of all deaths in Japan, including heart disease, stroke, cancer, and pneumonia. Each increased in the past year, with the exception of homicide and suicide. The category “other,” which is a collection of all other causes, rose 5.9%. The sharpest increases occurred immediately after the meltdowns, in March-June 2011 (vs. the same period 2010), a finding consistent with that found in preliminary mortality in the U.S. in a December 2011 article I co-authored with Dr. Janette Sherman in the International Journal of Health Services.
Well, the findings of Mangano and Sherman from last year are not worth the paper they are written on, so to claim consistency with the earlier cherry-picking that they cooked up does not strengthen his argument. Furthermore one can find other time periods where the increase in deaths are as high or higher than during the time period immediately after the meltdowns, for instance for the 4 month period June-September 2010, see the table below (marked in grey, while March-June 2011 is marked in green). Finally we see that there was an increase in the total number of deaths for 2010 that amounted to around 55000, i.e. the same order of increase as for 2011. Once again without any nuclear disaster or tsunami to blame. In the table below months marked in yellow has an increase larger than 4% with respect to the same month the year before. As seen there are many months with increases of the same order that the one Mangano wants us to believe is spectacular. The average increase for the years 2004-2011 is 2.8%, about twice the 1.5% that he considers to be normal.
What about the "other" deaths?
In the table above we have the increases for all deaths, let's look closer at the category "other" that Mangano claims increased with 5.9% during 2011. First a correction, Mangano says that the Japanese statistics reports provide mortality numbers for 12 common causes. He may have looked at some simplified table, there are about 130 different causes available in the tables I have looked into, all classified according to some standard (WHO?). What Mangano does not mention is that the remainder of unexplained causes in the category "other" amounts to around 5000 per year, thus reducing the number of "unexplained deaths" by a factor of ten. Furthermore, as seen the total amount in this category for 2011 is 5328, only about 150 more than in 2010.
Please note the relatively large jump in 2010 compared with 2009. Why be concerned with the increase of 150 (2.9%) for 2011 while ignoring the huge increase (13.1%) for 2010? If anybody wonders, Mangano stated an increase of 5.9% while the table below has only 2.9%, but he compares data from March of a given year until February the next year, while I use the average of the full year.
A question remains regarding the use of phrases "with no obvious cause" and "unexplained deaths". Out of the 37,800 deaths that Mangano claims are unexplained 28,800 of them are classified and identified, there are only 5,000 that lacks an explanation in the MHLW data base. That does not mean that they are mysterious in any way, they are cases that do not fit into any of the 130 categories. And if we are talking about health effects due to radioactivity, why look into this category? Wouldn't it make more sense to look at some of the 130 categories that could have some relation to exposure to radioactivity? Ah, silly me, that would require that Mangano has a mechanism to explain why the radioactivity would cause increased number of deaths so close in time after the nuclear meltdowns. Just like last year's cherry-picking exercise together with Janette Sherman he has no mechanism. It is in their interest to have us believe that radioactivity kills babies and adults in some mysterious way, then refer to other sources that claim health effects, like the Yablokov Chernobyl book that was edited by Sherman.
Let us cherry-pick a bit
Below is a plot with the monthly increase of deaths in the category "other", the way Mangano wants us to percieve it:
There is indeed a sharp increase the months after Fukushima. Now, let us add the other years from the table above:
Shame, Joe, shame.
Not that I want to give him any ideas, but If I was him I would rather count the absolute number of deaths, as shown below. The increase for the months March-June 2011 does not look as impressive as in the previous plot, but the peak is more difficult to discard when comparing with earlier years. On the other hand the problem remains that the increase is not very remarkable. And although some of those deaths could be related to all the things happening after one of the world's worst earthquakes ever, the tsunami and the chaotic and stressful time period with the nuclear meltdowns, Mangano still lacks a mechanism for why any of those deaths would be due to radioactivity. And he has used data for all of Japan, nice to see that he is too lazy to use the regional data and thereby try to correlate increases in deaths with the radioactive fallout. Maybe this is what he will do in the next article, can we guess that there will be one showing up in International Journal of Health Services within a couple of months?
Some final comments
The last section of Mangano's fearmongering text has two sentences.
There is no question that even if Fukushima studies proceed and are conducted in an objective manner, it will take years before the true extent of casualties are known.
Meanwhile we should live in fear by listening to Mangano's lousy kind of research? Why did he even bother writing this article and once again perform a cynical cherry-picking exercise if he acknowledges that it will take time to know the true extent of the disaster? It is also noteworthy how he seeds a conspiracy theory by writing "even if Fukushima studies [...] are conducted in an objective manner". Instead of trusting the corrup authorities and all the dishonest (real) researchers we should all listen to Fairiewinds or ENENews for the truth, or?
However, an early estimate of 38,700 additional unexplained deaths in Japan in just one year must be taken seriously, and underline the need for Fukushima health studies to be made a top priority, in Japan and in other affected nations.
No, it just shows that Joseph Mangano displays his dishonesty and incompetence in the field once more. If these deaths are unexplained, there are larger mysteries in the data from time periods without nuclear meltdowns. It should be emphasized that the number of unexplained deaths are about 5,300, not 38,700, and that is only 150 more than for 2010. We also note that the variations from year to year varies more than this number, irrespective of nuclear disasters. Fukushima health studies are indeed a top priority, and there may be some unexpected findings coming out from them. But health effects in epidemiological studies are difficult to interpret as it is, and therefore they are much better conducted without charlatans like Mangano, Busby and Caldicott adding their nonsense into the brew.
Mattias Lantz - member of the independent network Nuclear Power Yes Please
Appendix - Why should the number of deaths in Japan increase every year
It is all due to the demographics of Japan. The population is peaking at the moment:
An ageing population (Japan is the among the countries in the world with the longest average life span) together with a decrease in the number of born children gives the result that at the moment there are more people dying than being born, and the discrepancy increases for every year. At the present rate the population in Japan is expected to decrease from 128 million at present to about 87 million in 2060 (see Wikipedia or other source for more information).