Pretty obvious, but here’s a 2010 study.
The study, just published in the International Journal of Design and Nature and Ecodynamics, starts with a puzzle about racing sports: “More and more, the winning runners are black athletes, particularly of West African origin, and the winning swimmers are white. More and more, the world finalists in sprint are black and in swimming are white.”
The authors–Edward Jones of Howard University and Adrian Bejan and Jordan Charles of Duke University–attribute the two trends to a common factor: center of gravity. They explain:
Anthropometric measurements of large populations show that systematic differences exist among blacks, whites and Asians. The published evidence is MASSIVE: blacks have longer limbs than whites, and because blacks have longer legs and smaller circumferences (e.g. calves and arms), their center of mass is higher than that in other individuals of the same height. Asians and whites have longer torsos, therefore their centers of mass are lower.
These structural differences, they argue, generate differences in performance. Using equations about the physics of locomotion, they analyze racing as a process of falling forward. Based on this analysis, they conclude that having a higher center of body mass in a standing position is advantageous in running but disadvantageous in swimming.
Drawing on data from 17 groups of soldiers around the world, the authors note that in terms of upper body length, “the measurements of the group of blacks fall well below those of the other groups. Their average sitting height (87.5 cm) is 3 cm shorter than the average sitting height of the group of men with the same average height (172 cm).” From this, they calculate that “the dimension that dictates the speed in running (L1) is 3.7 percent greater in blacks than in whites. At the same time, the dimension that governs speed in swimming is 3.5 percent greater in whites than in blacks.”
Measurements of women suggest a similar pattern:
[U]pper- and lower-extremity bone lengths are significantly longer in adult black females than in white females. For the lower-extremity bone lengths, the difference is between 80.3 ± 10.4 cm (black females) and 78.1 ± 6.2 cm (white females). This difference of 2.2 cm represents 2.7 percent of the lower-extremity length, and it is of the same order as the 3.7 percent difference between the sitting heights of whites and blacks.
The paper calculates that a 3 percent difference in center of mass–the average difference between blacks and whites–produces for the athlete with the higher center of mass
a 1.5 percent increase in the winning speed for the 100 [meter] dash. This represents a 1.5 percent decrease in the winning time, for example, a drop from 10 to 9.85 [seconds]. This change is enormous in comparison with the incremental decreases that differentiate between world records from year to year. In fact, the 0.15[-second] decrease corresponds to the evolution of the speed records … from 1960 (Armin Hary) to 1991 (Carl Lewis). The 3 percent difference in L1 between groups represents an enormous advantage for black athletes.
For swimming, the conclusion is quantitatively the same, but in favor of white athletes. The 3 percent increase in [lower-body length] means a 1.5 percent increase in winning speed, and a 1.5 percent decrease in winning time. Because the winning times for 100[-meter] free are of the order of 50 [seconds], this represents a decrease of the order of 0.75 [seconds] in the winning time. This is a significant advantage for white swimmers, because it corresponds to evolution of the records over 10 years, for example, from 1976 (James Montgomery) to 1985 (Matt Biondi).