Here we go again. Nikolay Lamm has come up with a new visualization of “average” people, except this time, where curvier women’s bodies were good, he clearly thinks that the fluffier American males in this exercise are doin’ it wrong. That is, assuming that you accept his weight distributions again.* (*You shouldn’t.) Other problems are as follows:
1. “Averages,” meaning means in Lamm’s case, are useless in the real world.
As I mentioned in my inaugural post, a collection of means does not a “normal” or real human make. Median would be a much better choice for this representation because it discounts the outliers.
2. Why the hell are you creating an “average” man based solely on BMI, waist measurement, and height?
Do I believe the “average” American man is fatter than the average Dutch, French, or Japanese guy? Sure. We have bad work-life balance, terrible commutes, an oversupply of cheap calorie-rich foods, and a bunch of other factors propelling us in that direction, and I do not deny that there are some serious health issues in America that are related to weight. HOWEVER, Lamm is not proving anything about the “average” American male. Without a chest, arm, neck, etc., measurement or percent body fat factored in, the numbers could be skewed because of the popularity of bodybuilding in America, for all we know.
BMI might be useful in a very broad sense, but trying to use it to bolster something like this with a straight face is like trying to use a backhoe to do the reaping in your backyard garden: missing the point. On each of the male figures, the shoulders and arms are very lanky, if not outright frail looking, and their stomachs distend to varying degrees below the waist. There is no hip measurement, so there is absolutely no evidence, as justified by Lamm, that the lower abdomen is where the weight is going. Hell, maybe there are mostly men with spindly arms in France and Japan, but lots of men with beaucoup upper body strength in the Netherlands and the USA.
Lamm tries to address what I can only assume is ongoing criticism of his work by people with more sway than I by saying that he consulted Dr. Matthew Reed, whose [pretty cool] research is of the exact type that works with idealized averages and not the range of realities. (For instance, in helping to design seats for public transit, Dr. Reed is not going to worry about the 600 lb commuter because it isn’t reasonable to make all seats big enough to hold the largest or smallest person.) I am not convinced that Reed advised Lamm to make their lower abdomens and derrieres larger because that’s how people actually look, in his expert opinion. Furthermore, most of Reed’s research focuses on ass- and other forms of spread with regard to seated positions, so if he did sign off on the final products as definitely representative of the way in which weight would be distributed on a male of the height, waist, and BMI measurements in question, I’d have to ask whether that’s because he spends so much time thinking about how much room people take up when seated. Hell, as a frequent rider of public transit and occasional airplane passenger, I think he and his colleagues are underestimating average ass spread.
3. You do not create a man of “average” size in a country with as many different races as ours independent of measurements by race.
Lamm sort of allows for this with a Jim Sciutto quote about “less privileged groups across all races” bringing the average height down, but he doesn’t specifically say that, in rendering a visibly Caucasian male, he used the white guy measurements. There are plenty of overweight white guys, too, don’t get me wrong, but the “average” white guy (which, I realize, is an especially useless way to view socioeconomic status given that most of our plutocrats are white guys and Lamm is using means, not medians) is better off than the “average” Black or Latino guy and income levels affect health and weight in undeniable ways. If we’re measuring a bunch of naturalized citizens who grew up in Mexico, for instance, it is very likely that they will bring the average height down, but raise the average waist and BMI measurements.
Similarly, though white guys outnumber males of other races in the US, there’s no telling how the inclusion of means from other races would skew the representation of an “average” white male. Certainly, this will be less problematic than if Lamm had tried to represent the “average” Latino male in the US using these numbers, but the point remains. Due to the presence of evil-otic (evil + idiotic) asshats in Congress, I can’t get the statistics I want on the races of adult males for this post, but according to Wikipedia, 12.6% of the US population is African American; 16.4% is Hispanic/Latino, and 4.8% is Asian American. We can reasonably assume that men comprise half of those totals, but I can’t factor out minors, unfortunately. This means that approximately 25% of US males are not white, yet Lamm lumps their genetic proclivities and other considerations in with the “average,” clearly white male.
Below is a table of the types of measurements Lamm used, but for males ages 20-39 by race:
Really, anything other than the measurements within each race are useless for the purposes of a simulation that appears to specifically depict a man of only one race.
I think this project could actually have been much more powerful if he had compared the “average” white, Black, and Latino males in the US (since race is the best proxy for income level in this CDC data) to those of the averages for different races in other countries because healthcare access and the lower cost of nutritionally void calories in the US is absolutely a problem. If the Latinos here are less fit than the Latinos there, for instance, it’s not because of genetics.
4. On the topic of race and genetics, Lamm does not seem to take genetics into consideration at all in his critique.
The chart above clearly shows that Black males ages 20-39 are taller and have smaller waists than do white men of the same ages, in spite of any income discrepancies that may have affected diet. Their mean BMI is also higher (as are the mean weights: 197.4 lbs for white males 20-39, 198.1 lbs for Black males), which could just be a symptom of the inherent weirdness of using all these means as though they scale in an idealized manner (e.g. assuming that “average” tall people have wider waists and weigh more than “average” short people), but it could also mean that the Black males are more muscular. Mexican-Americans, on the other hand, seem to be distinctly shorter, wider, and heavier than their non-Mexican Latino counterparts.
Clearly, genetics cannot be ignored when considering the varying degrees of fitness of different populations. Lamm, however, does just that by including all races’ measurements in his “average” American male and then mentioning things like diet and healthcare in passing as though they affect the white and non-white male populations of the US equally (which is the only way to justify having a white model but blaming his comparative heft on these factors). All of this in line with the Sciutto quote about “less-privileged groups” and Lamm’s message seems to be, “This guy is fat in part because people who don’t look like him are medically and nutritionally disadvantaged.” I guess that’s… not inaccurate… but then why not make the model mixed race?!
This criticism is not limited to his American model, but I am better able to comment on the complications that arise in homogenizing US males. For the sake of comparison, Lamm’s other countries of choice by immigrant population:
The Netherlands: 81.7% white, unable to find a breakdown of race by age so as to discount those under ~20 who are the first-generation children of immigrants.
France: “As of 2004, French think-tank Institut Montaigne estimated that there were 51 million (85%) white people or European origin, 6 million (10%) North African people, 2 million (3.5%) Black people and 1 million (1.5%) people of Asian origin in Metropolitan France, including all generations of immigrant descendants.”
Japan: “98.5% Japanese and 1.5% other. The concept of the ethnic groups by the Japanese statistics is different from the ethnicity census of North American, Australasian, Brazilian or some Western European statistics. For example, the United Kingdom Census asks ethnic or racial background which composites the population of the United Kingdom, regardless of their nationalities. The Japanese Statistics Bureau, however, does not have this question yet. Since the Japanese population census asks the people’s nationality rather than their ethnic background, naturalized Japanese citizens and Japanese nationals with multi-ethnic background are considered to be ethnically Japanese in the population census of Japan.“
While I understand the appeal of using easy-to-visualize averages to propel social commentary, I don’t feel that it’s contributing to an informed discussion of the underlying problems. We already know America has a weight problem. If anything, this probably makes the issue seem less dire because Lamm’s “average” American male doesn’t have far to go if he wants to be healthier. In my opinion, the assumptions in conversations like this need to be made explicit, the data needs to be more robust, and the limitations need to be respected. Otherwise, you end up with a confused social commentary like Lamm’s which criticizes the everyman for being overweight, but glorifies the “normal” woman using the same data sets. Presented merely as art, it’s a good starting point for a more serious conversation, but Lamm clearly has an agenda of his own (which is fine) that The Internet has too readily accepted on its feeble scientific merits.