Muehlenkamp’s body-mass delusion: a case of BMI abuse

Foundational to Roberto Muehlenkamp’s misrepresentations of the realities concerning burial space at the Reinhardt camps is his claim that Polish Jews weighed an average of 34 kilograms (75 pounds). This claim is based on the assumptions that (1) one third of the Jews were children (meaning under the age of 15), that (2) the adults weighed an average of 43 kg, and that (3) the children weighed an average of 16 kg.

This post will show how worthless are Muehlenkamp’s claims, and how mendacious his analysis. The passage under discussion is from Muehlenkamp’s contribution to the holocaust controversies manifesto, pp. 416-418, especially pp. 417-418; the results of this vital passage exert themselves throughout the rest of Muehlenkamp’s discussion of both burial space and cremation.

Determination of the average height

Muehlenkamp begins by arguing that the Jews of Poland were quite short. For sources on the height of Jews, Muehlenkamp cites Provan’s article – the same one he lied about – without making any effort to check the sources. Provan gives three sources as follows:

according to ethnological studies done by Dr. Otto Von Verschuer, the Jews of Poland were about three inches shorter than the average German. This comparative smallness is confirmed by other authorities, notably John R. Baker and Lothrop Stoddard.

I will address the two authorities confirming the Jews’ smallness first. In his 1974 book, speaking of Ashkenazi Jews, Baker observes that “The legs are rather short, and the total height is therefore moderate or rather short.” There is no reference to heights at during specific time period, nor to the Jews of Poland in particular, nor are any numbers given for the height of this population. Moreover, Baker’s attribution of short legs to Ashkenazi Jews would, if accepted as true, tend to undermine Muehlenkamp’s belief that the Polish Jews had a particularly low average BMI. At a given height, individuals with long legs will tend to be lighter than individuals with short legs (and correspondingly larger torsos); long-legged populations will tend to have lower average weights than stubby-legged populations at the same height. Indeed, Baker observes that “The shoulders are broad and the form is thick-set” – not a build associated with a low BMI.

Stoddard’s 1924 book discusses the Jews of Poland, Russia, and Romania, which it terms Ashkenazim, and which it contrasts with the Jews of western Europe, which it terms Sephardim (the distinction does not precisely correspond to the current use of these terms). According to Stoddard, the Ashkenazim “possess very little of the old Semitic Hebrew blood”. Rather, their genetics are heavily influenced by the Khazars, and it is from this source that the “dwarfish stature, flat faces, high cheekbones, and other Mongoloid traits so common among east-European Jews” derive. Does Muehlenkamp endorse this analysis? (The matter is ultimately irrelevant, as Stoddard offers no numbers for height, rendering his analysis useless for our present purpose.)

Now, let’s consider von Verschuer, who Provan claims carried out ethnological studies measuring the difference between the heights of Jews and Germans. One wonders whether Muehlenkamp is even aware of the identity of the scientist to whom he is appealing here. Probably not, as neither he nor Provan manages to give the man’s name correctly. Otmar von Verschuer was in fact Joseph Mengele’s supervisor, and the source of Mengele’s interest in twins. His article cited by Provan, a contribution to the series Forschungen zur Judenfrage (does Muehlenkamp also endorse the other claims made in that series, or even all those in von Verschuer’s article?) does compare the heights of Jews and Germans, but contrary to Provan’s claim there is no indication that von Verschuer carried out the studies himself. Unfortunately, the source for the statements on the height of Jews is not indicated very clearly (the source for the height data for Germans, on the other hand, was fairly up-to-date at the time, at least if we can go by publication dates – but without checking, who knows when the original studies were done), but it seems that the data for Jews and for Germans may derive from different studies, which raises the question of whether the data are really comparable, or whether (for example) the data for Jews might be older than that for Germans, which could account for some of the difference in height (given the increase in average human height over the relevant time period). Von Verschuer was not a specialist in the topic, and merely mentioned it in passing, so it would be no surprise if, the increase in human height over time being less well known at that time than it is today, and there being a certain ideological pressure to find Jews to be inferior to Germans in height, he did not consider all factors that could influence his comparison. (In support of this possibility, one can mention the fact that the 1906 Jewish encyclopedia claims that at that time Jews were only moderately – say, one inch – shorter than the non-Jewish populations among which they lived. The precise height data offered therein is naturally unusable, though, given the general increase in height over time.) Moreover, contrary to Provan and Muehlenkamp’s claims, von Verschuer makes no reference to Polish Jews specifically, but speaks of Jews in general.

Muehlenkamp has recently promoted von Verschuer’s work (in reality not his work, but a claim which he sources from other authors) to the status of contemporary studies, stating that the Jews of Poland were “of comparatively small stature according to contemporary anthropological studies (average height 1.60 meters)”. First of all, the value 1.60 meters is Muehlencamp’s own inference and not something he found in any other source. Second, not one of the sources cited by Provan and appropriated (without checking) by Muehlenkamp says anything about the height of Polish Jews in particular. Third, the publication dates of the works which von Verschuer cites range from 1883 to 1938. As Muehlenkamp has made no attempt to trace the matter back to the original source, what basis does he have for the claim that these are contemporary (or even near-contemporary) studies?

As I have not made any serious attempt to investigate the height of Polish Jews during the 1940s, I will refrain from speculating on how accurate Muehlenkamp’s assumptions regarding their height may be. What I can say, however, is that Muehlenkamp has not done anywhere near the necessary research to make confident claims about this issue.

Determination of average BMI

BMI, or body mass index, is a tool meant to give a rough idea of what one should weigh at a given height. It is not particularly precise, and does not work well for all populations. As in indicator of its highly approximate nature, one need only consider that it gives the same suggested weights for both men and women of the same height.

Muehlenkamp refers to a website (though he fails to give the url) for a claim that underweight individuals have a BMI between 15 and 18.8. Based on this, he gives the figures of 38 and 48 kg (corresponding to BMIs of 14.84 and 18.75 at a height of 1.60 meters), and takes their mean, 43 kg. This corresponds to a BMI of 16.8. Note that these numbers are pure inventions on Muehlenkamp’s part, and rest on no data whatsoever.

As a preliminary source, let’s consider the international data available at the wikipedia entry for BMI. Observe that the country with the lowest average BMI is Eritrea, where the average BMI is 19.85. If we consider females only, then the lowest average BMI, again from Eritrea, is 19.43. In no country does the average BMI come anywhere near Muehlenkamp’s supposed Jewish BMI of 16.8, despite the fact that there are many countries where people are poorly fed and have body types decidedly more slight in torso and long in limb than those of Ashkenazi Jews.

Still, Muehlenkamp may claim that the Jews of Poland were starved to a special extent, which more than counteracted the effect of their comparatively short-legged, stocky builds. This site can help us in our search for starved populations with which to compare. It allows us to search for data on the percentages of certain populations falling in the <=17 and <18.5 BMI ranges. A glance at the data (look under "Detailed Data", and be forewarned that there are a few weird outliers that are wildly different from the surrounding data, which suggests an error in data entry) shows that the percentage in the <=17 BMI category is consistently fairly low, even in studies that examined only females (and even in worse-nourished sub-populations, such as uneducated females of low socioeconomic status, which were separated out). In the <18.5 category, however, while the percentages are generally low, there are a few cases where the numbers are considerably higher, and in certain segments of the female population even reach a moderate majority. Thus in the worst-nourished regions of the world, disadvantaged females may have had a median (not mean, which is what we want; moreover one would expect the mean to exceed the median here) BMI of slightly under 18.5. How, precisely, can Muehlenkamp justify his claim that Polish Jews of both sexes had an average BMI of 16.8? What data, exactly, support this claim?

Determination of average weight given average height and average BMI.

Without feeling the need to spell anything out, Muehlenkamp assumes that the mean weight of a population can be calculated as the weight of an individual whose height is the mean height of the population and whose BMI is the mean BMI of the population. That is, Muehlenkamp assumes the truth of the following proposition:

PROPOSITION (Muehlenkamp): Given a population P, the mean weight W_mean(P) of members of P can be calculated via the formula W_mean(P) = the weight of an individual with height h and BMI c, where c is the mean BMI of the members of P and h is the mean height of members of P.

In fact, the above proposition is false. The underlying reason that the proposition is false is that the function W_c(h) sending a height h to the weight of an individual with height h and given BMI c is not a linear function but a quadratic function. The details should be instantly obvious, and the error mortifyingly primitive, to anyone with the slightest business using numbers to analyse anything more complex than his household budget (a group of people which does not include Muehlenkamp).

The above implies that Muehlenkamp’s method for obtaining data on average weights given only data (or assumptions) on average height and data (or assumptions) on average BMI cannot work.

The average weight of the children

Muehlenkamp’s given average weight of 16 kg for Jews under age 15 is the result of rounding up the figure 15.6 kg, which in turn comes from the assumption that average adult weight should be 2.76 times average weight for the under 15, and the assumption (falsely derived, as shown above, and based on unreasonable suppositions) that the average adult weight was 43 kg. Let’s look at some data on the weight of children from the age of 2 upward. Naturally, I am not suggesting that the population represented by this data has the same mean weight as the Polish Jews. Rather, this gives us an actual population from which to begin, in order to impose some degree of realism on this analysis. Let us suppose that the average Polish Jew under age 15 had a weight equal to the 3rd percentile in the above linked tables. In fact, to skew things even more in Muehlenkamp’s favor, let’s assume that it corresponds to the 3rd percentile weight for females.

We begin by summing the 3rd percentile weights for females from months 24.5 through 179.5. The above data is missing information for weights at age less than two years. To fill the gap, consider this data for children up to the age of 2. While it does not have the data for various percentiles, we can sum the data for each month (0-23), then divide by the value for month 24. The result is that the sum of the weights at months 0 to 23 amount to 17.4 months’ worth of weight at age 24 months. Plugging in the 3rd percentile female weight at 24 months, adding this to the total obtained above, and dividing by 180 (the number of months we added, i.e. the number of months under age 15) we obtain an average weight of 20.15 kg, dramatically more than Muehlenkamp’s figure of 16 kg.

A slight error has been introduced here in that we have supplemented our mid-month data for ages above two years with beginning of month data for the first two years, but the error works in Muehlenkamp’s favor, and therefore causes us no problems. One should add that this is all based on the assumption that the Jews under age 15 were uniformly distributed in age, and that (as explained below) this causes an underestimate of the actual average weight, because in reality there were comparatively few at the lowest ages.

Granted, this is not based on specific data for Polish Jews in the 1940s, but neither is Muehlenkamp’s work. The assumption that the young Jews’ average weight was equal to the 3rd percentile of the female weights in the data linked above was really quite generous. What evidence does Muehlenkamp have that their weight was lower than this? Or even that it was this low?

The proportion of the population which were children

Basing himself on Carlo Mattogno’s reference to a study claiming that in 1931 29.6% of the Polish Jews were under age 15, Muehlenkamp assumes that he can calculate average weights on the basis of the assumption that one out of three Jews was under age 15. First of all, the rounding of 29.6% up to one third was an approximation which Mattogno was entitled to make because it worked against his argument, but which Muehlenkamp is not equally entitled to make because it introduces error in favor of his argument.

More important, however, is the fact that demographics in 1931 do not guarantee anything about demographics over a decade later. Given that Poland experienced Jewish emigration during the 1930s, which can be expected to have come predominantly from the younger age groups and have had the effect of decreasing the birth rate, there is reason to suspect that, even if the figure of 29.6% for 1931 is correct, the figure in 1939 was lower than this. Moreover, Jewish birth rates are reported to have plummeted during the war. The recorded information concerning birth rates in the Warsaw ghetto, for instance, shows a vanishingly small number of births. Evidently this would have had a significant impact on the proportion of the population under the age of 15. Moreover, it would imply that the under-15 population would skew towards the older part of the 0-15 range, which implies that the average weight would be higher than it would be on the assumption that there were equal numbers at each age. Needless to say, Muehlenkamp simply ignores these factors.


Muehlenkamp’s determination of the average weight of Polish Jews, which forms the foundation of his entire contribution to the manifesto, is worthless. It is based on an ignorance of his own sources, on unjustified and unjustifiable numerical assumptions, and on mathematical errors. As the results of this attempted determination are used throughout his analysis of both burial space and cremation, the errors and absurdities in Muehlenkamp’s determination of the average weight invalidate all the rest of his calculations.

This entry was posted in Uncategorized and tagged , , , , . Bookmark the permalink.

3 Responses to Muehlenkamp’s body-mass delusion: a case of BMI abuse

  1. Pingback: Taking out the garbage | Holocaust History Channel

  2. Pingback: Pearls before swine: explaining the obvious to an idiot | Holocaust History Channel

  3. Pingback: Another act from Muehlenkamp’s comedy of errors: 50 kg of wood has the same energy content as 12 kg of wood | Holocaust History Channel

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s