A Size 2 Is a Size 2 Is a Size 8
Why clothing sizes make no sense.
Illustration by Mark Alan Stamaty.
If you are like me, you may have had the following deflating experience: After confidently selecting a pair of jeans in your size from the rack, you find yourself alone in a tiny dressing room, desperately in need of heavy machinery to get them onto your body, wondering why, oh, why clothing sizes don’t make sense.
Cheer up! You’re in excellent historical company; people have been moaning about sizes for the better part of a century. “I don’t know who the mythical size 36 is who forms the basis of sizing,” complained one indignant retail executive in a 1927 New York Times article, “but average, tall, short, thin and plump women come into a department store and the 36 size fits none of them.” A young Katharine Graham reported for the Washington Post on a similar subject in 1948: “As most harassed mothers know,” Graham wrote, “size 5 in a little girl’s dress can mean almost anything.” It’s enough to make you wonder: Was there ever a time when sizes were standardized?
The federal government did take a stab at it in the early 1940s. That was when the Depression-era Works Progress Administration commissioned a study of the American female body, an effort to instill a method to the sizing madness. At that point, the ready-made clothing industry was in its infancy. If your clothes are made-to-measure—as they were in an earlier era, particularly for wealthy women—there’s no need for a standard set of sizes. But as European couturiers were hobbled by World War II, an American fashion industry developed, with New York as its capital city. New York’s all-star cast of designers, among them Claire McCardell and “Sophie of Saks,” specialized not in couture but in ready-to-wear. And as wealthy women began to purchase premade clothing, pressure mounted to ensure that it fit in a consistent way.
The ready-to-wear sizing system that existed prior to the ‘40s was first developed for menswear. Scholars have found evidence of standardized men’s sizing as far back as the Revolutionary War. By the War of 1812, the Army was in the practice of holding stocks of ready-made uniforms sized according to a single measurement, of the chest—based on the assumption that you could deduce from it a proportional understanding of the rest of a man’s body. So, when manufacturers in the early 20th century began to produce women’s clothing, they based women’s sizes exclusively on a single measurement: the bust.
The only problem? Bust measurements on their own are not particularly accurate indicators of a woman’s size or of the rest of her proportions. As we all know, some small women have very large breasts, and some large women have very small ones. This sizing conundrum was particularly irksome to the Mail Order Association of America, which was well aware by the late 1930s that women often returned clothing because of poor fit.
In an attempt to resolve this problem, the Works Progress Administration gave two statisticians, Ruth O’Brien and William Shelton, a grant to conduct a nationwide survey of women’s measurements. O’Brien and Shelton sent more than 100 trained operatives out into the field, where they measured almost 15,000 women, taking 59 distinct measurements of the female form. When the data came in, O’Brien and Shelton hoped to determine whether any proportional relationships existed among measurements that could be broadly applied to create a simple, standardized system of sizing.
Sadly, the data didn’t cooperate. The best system O’Brien and Shelton proposed was based primarily on height and weight. But the statisticians wisely recognized that women might balk if a sizing system forced them to confront—or tell a sales clerk—their own exact poundage. As an inferior alternative, they imagined a system based on a single measurement of the upper body and combined that with a height index (regulars, longs, and shorts), and a lower body girth index (regulars, stouts, and slims). “The nomenclature adopted would probably have to be on some arbitrary basis, as is the present system of shoe sizes, rather than representing the numerical values of some anthropometrical measurement,” O’Brien and Shelton wrote, summing up what is perhaps the pair’s most lasting contribution to modern sizing: Even though a measurement can be extrapolated from a size, the actual numbers we use in women’s sizing are not, themselves, measurements.
“It was pretty impressive, what they did,” says sizing scholar Lynn Boorady, noting that they devised something like 27 different sizes. But, she says, “It was an obscene number, and obviously not useful for anyone manufacturing.” It wasn’t until almost a decade later that the National Bureau of Standards* (now the National Institute of Standards and Technology) reanalyzed the O’Brien/Shelton data and came up with an official system—one that has served as the basis for all future systems. Women’s sizes were derived from bust size—with all other measurements based on the proportions of an hourglass figure—and represented by even numbers from 8 to 38. These basic sizes were combined with a T, R, or S, to indicate height, and a plus or minus, to represent lower body girth. The system was published as a commercial standard—a recommendation, legally required only for the pattern-making industry—in 1958. The industry was at first enthusiastic about these recommendations, and major mail order companies like Sears and Montgomery Ward started to adopt these government sizes. But enthusiasm eventually flagged. By 1970, the NBS downgraded its Commercial Standard to a Voluntary Product Standard, and by 1983, the government withdrew the standard entirely, damning future generations to inconsistent fits and many, many mail order returns.
Why did we choose chaos over the government’s system? To begin with, the O’Brien/Shelton data was flawed. The study accounted only for white women; women of color who came in were measured, but their measurements were discarded. And since the study offered a small stipend to anyone who volunteered to be measured, there’s a decent chance that the results skewed toward the poor and malnourished. When the NBS re-analyzed the data to produce the commercial standard, they distorted results even further by adding the measurements of women who had served in the Army during World War II—likely among the most fit women in the population.
Julia Felsenthal is an assistant at Slate.