But what if manspreading — the act of a male passenger sitting on public transit with his legs apart — is not simply rude behavior or chauvinism?
Our new analysis suggests that manspreading is something men do to adjust for their body proportions — especially their high shoulder to hip ratio — and not an act of transgression against their fellow passengers.
In fancy science-y language: Based on our multivariate analysis of anthropometric parameters across multiple data sets, manspreading appears to be an adaptive strategy that men employ due to innate morphological characteristics.
One of the data sets that we studied show that the average man’s shoulders were far wider — 28% wider, in fact, than his hips. By contrast, the average woman’s shoulders were only 3%[†] wider than her hips.[‡] 
Take a look at the cartoon man in the MTA campaign to “shame” manspreaders. His body proportions are all wrong. The cartoon body does not resemble the carbon based men that we’ve described in the data above — who have shoulders that are nearly a third wider than their waists. (We’ve added green bars to demonstrate how the MTA’s campaign distorts men’s bodies — showing their shoulders and waists at the same width — like some bizarre, dysmorphic Barbie doll.)
Why do body dimensions matter so much?
A bench seat on the subway is basically two planes that meet at a 90-degree angle: So, if a man sits on the subway with his knees together, and other passengers crowd in closely on both sides, then his torso likely won’t fit on the top half of the seat if his knees are positioned less than shoulder width apart.
In a sense, a man’s knees may function like the proverbial cat’s whiskers — used as a kind of heuristic to test if there is ample space for his torso to fit on the seat.
A second data series reported measurements from military personnel. These measurements were taken slightly differently but yielded similar conclusions overall. The measurement captured forearm-to-forearm breadth, which may offer additional insight into the manspreading phenomenon. The military data reported almost equal differentials between men and women’s upper torso breadth. But, even more interesting, in this data set measuring close to 4,000 people, women and men were found to have very similar distributions of hip breadths. This suggests that, proportionally, a man needs to secure more seat space using his legs than a woman would need to in order for the man to maintain enough room to sit up straight in his seat. 
(Both co-authors of this post report being forced to hunch over awkwardly while riding on crowded subway trains: It’s anecdotal — but it’s still a really annoying experience.)
Finally, an adaptive benefit for manspreading may be to avoid collisions in the aisles on crowded trains. Applying calculations to a third data set, Kodak’s Ergonomic design study, demonstrates that a seated man’s knees extend forward much farther than a seated woman’s knees do. On average, a seated man’s knees are 0.8 inches longer than a seated woman’s — but can extend out as much as 4.3 inches longer (95th male percentile, 5th female percentile) than a seated woman’s knees would. The average man can reduce the distance which his knees protrude by about 3.1 inches[§] simply by adjusting his legs outward — manspreading — to a 30-degree angle. 
Even if the OED says that Manspreading is a real word — it may not be a real thing.
Mark Skinner trained as a data scientist and works as a Country Insights Analyst at Roubini Global Economics.
[†]Data Table 1: Hip and shoulder breadth measurement and calculated percentages.
[‡] To account for variations, we cross referenced extreme cases from the data set and found that even in instances where a man is in the 95th percentile waist breadth and in the 5th percentile shoulder breadth that man would still have shoulders that were a full 8% wider than his hips, if such a man exists.
[*] Original analysis by Skinner and Bennington. (Based on data from Gordon C. et. al 1988.)
[§] We calculated this value by taking the average femur length for men and subtracting the √3 (proportioned) leg of the triangle formed with his femur as the hypotenuse of a 30-60-90 triangle. i.e. avg. femur length * (1-√3/2).