It's pretty strange to think that the option of buying clothes off the rack is really a very recent invention. Up to the 1920's, almost all women's clothing was made at home or by dressmakers based on an individual's body measurements.
It wasn't until the 1920's that a woman could walk in a store, see a dress she liked, buy it, and bring it home that day. After centuries, nay millenia, of having plenty of time to consider a new dress, consumers now had instant gratification. This led to new psychological phenomena like shop lifting and retail therapy.
Another dark side of ready to wear also emerged: standardized women's measurements.
In the early years of mass-produced ready to wear, sizing was chaotic because there were no standards for manufacturers to follow. Mail order companies had to deal with masses of returns and consumers had to sort through differing sizes produced by different manufacturers. At the request of manufacturers, the US Department of Agriculture undertook a large scale project in 1937 to measure 15,000 women volunteers and statistically analyze the data with a goal to make recommendations for standard sizes. The project was partially funded by the WPA to provide employment during the depression. The results were published in 1941 and can be read here.
Women volunteered through recruiting organizations to be measured in 59 places wearing only a "measuring costume" consisting of knit shorts, paper shoes and a "bandeaux" and allowing themselves to be written on with ink. At least they did receive a small fee for participating in the study.
But, the fact that this study took place during the depression, and that it was a source of income may have skewed the data toward underweight body types. In addition, the study was explicitly created for white women-- measurements for women determined to be "not of the Caucasian race" were sometimes taken "for the sake of good feeling" within an organization--but this was noted so the results could later be rejected.
It also became apparent during the study that too many young women (perhaps with positive body image and spare time?) were self-selecting for the study, thus making the the data more youthful. In addition, the study over represented non-working, single, urban women. The Midwest and Rocky Mountain States--my places-- were not represented at all. The researchers also noted that rural women were consistently larger than urban women, yet they were also underrepresented.
Still the study went on and the measurements were statistically "corrected" by eliminating about 5,000 women's hard-earned numbers. Finally! the Average American White woman was discovered. Her measurements? Bust: 35.6 (90.4 cm), waist: 29.15 (74 cm), hips: 38.82 (98.6 cm).
But an average of all women is less important than correlations. As women get larger, the busts and waist measurements often increase more than the hip measurements. One thing that seems strange from our perspective is that an underbust measurement was not taken, so it was possible to study the correlation of women's bust to arm measurement, or height to bust, but not to study cup size in relation to anything. The most common bust to hip correlation in the study was bust of 33" (84 cm) and hips of 37" (94 cm) (a difference of 3"(7.5 cm) ) with a height of 5'3"inches (160cm). Ultimately, this study produced a wealth of information but no standard sizes. That's right, after undertaking all that measuring and computation and compiling of pages and pages of charts the authors concluded:
"From the point of view of the clothing designer, however, the measurements of the average woman are of limited usefulness. The Nation's women vary far too much in size to be properly fitted by garments made for the average woman."
There went the taxpayer's money, again.
So pattern companies seem to have arrived at their own numbers (does anyone reading this know how?) I do have a copy of DuBarry Patterns size charts from 1940. At this time, their "average figure," had a height of 5'3"-5'6" (160-167 cm) and "hips 3 inches larger than the bust and hips 9 inches larger than the waist." So in some ways (the 3 inch difference between hips and bust) the patterns did reflect the USDA's results. These measurements are certainly vastly different than those used in the patternmaking books of the early 20th Century that often had a 36" bust correlated to a 24" waist, probably because the women were no longer wearing corsets in the 1930s.
Because the 1937 study did not make any specific sizing recommendations, the Mail Order Association of America asked in the mid 1940s that the data be tranferred to the National Bureau of Standards to be re-analyzed. In order to have a larger sample, the measurements of 6,510 US WWII WACs (AIr Force Women) were added. If anything, this data must have further pushed the measurements towards a slimmer standard.
Another huge difference was that the new standards called for measurements to be adjusted to "compensate for the effect of foundation garments" " i.e a brassiere and girdle." So instead of the more natural "measuring costume" of 1937 seen above, the subjects of the later study were expected to wear 1950's style underwear:
Finally, in 1958 the sizing was published as Commercial Standard 215-58. Although I can't find a copy of these standards, sewing patterns give a pretty good indication of what was happening. While chest measurements remained constant, the standard now had hips at only 2" (5cm) larger than the bust,
The standards were revised and re-published in 1971, but the baseline data remained those same white, urban women from the depression era. And while height, bust and hip sizes had to grow to accomodate taller and larger women, waistlines often stayed the same. From the 1930's to the 1970s, the ideal became taller with a larger bust in relation to waist and hips. Look at the bottom of the page to see these sizing changes. Although women's lower halves were not shrinking in relation to their busts, new wonder fabrics were sculpting the female form.
Currently, most manufacturers have stopped using the government's standards and instead create their own size guidelines and then find "fit models" to design clothes on, anticipating the desired measurements of their desired niche market. According to one source (that I can no longer find) only sewing pattern companies continue to use the old standards--and from what I have experienced, the fit often has little to do with the sizing chart.
So here we are in the contemporary period of mass-produced ready to wear, sizing is chaotic because there are too many standards for manufacturers to follow. Internet retailers have to deal with masses of returns and consumers have to sort through sizes produced by different manufacturers that fit differently.
Is it time to go back, way back ?
Or perhaps into the future, with patterns generated from computer scans and clothing made to measure by 3-D printers ? Or somewhere in between, like this sewist?
The birth of 36-24-36 as a feminine ideal?
Sewing pattern sizing charts Source
1939-1958 (1940 Du Barry has a larger waist in smaller sizes)
SIZE 10 11 12 13 14 15 16 18 20 40
BUST 28 29 30 31 32 33 34 36 38 40
WAIST 23 23.5 24 25 26 27 28 30 32 34
HIP 31 32 33 34 35 36 37 39 41 43
SIZE 9 10 11 12 13 14 15 16 18 20
BUST 30.5 31 31.5 32 33 34 35 36 38 40
WAIST 23.5 24 24.5 25 25.5 26 27 28 30 32
HIP 32.5 33 33.5 34 35 36 37 38 40 42
1968 to present
SIZE 4 6 8 10 12 14 1 6 18 20 22 24
BUST 29.5 30.5 31.5 32.5 34 36 38 40 42 44 46
WAIST 22 23 24 25 26.5 28 30 32 34 37 39
HIP 31.5 32.5 33.5 34.5 36 38 40 42 44 46 48