Back to Trend Reports

Trend Reports | 7 min read

The Twenty-Year Rule Does Not Match the Measurements

The two people who actually put numbers on the fashion cycle measured a wave close to a century, and neither of them arrived anywhere near twenty years.

The Twenty-Year Rule Does Not Match the Measurements visual notes
Trend Reports notes from Iris Caldwell.

The claim arrives every season with the confidence of a physical law. Fashion returns every twenty years, so keep anything long enough and it will come back around. Nobody ever attaches a source to it. Two people did try to put numbers on the question, one by measuring three centuries of dresses and one by writing down how long a garment takes to stop looking embarrassing, and neither of them landed anywhere near twenty years.

This is not a pedantic quarrel. The twenty-year rule is the reason people keep clothes they will never wear again and buy revivals at the exact moment those revivals are about to date. Replacing it with the numbers that actually exist changes both decisions.

Where the twenty-year figure comes from

The appeal is obvious once you say it out loud. Twenty years is roughly the distance between a teenager and their parents, so the clothes that feel newly strange to one generation are the clothes the previous one wore in public. Retail likes the number too, because it turns any archive into a schedule and gives a marketing team a reason to reissue something.

What the figure lacks is a study behind it. It circulates as folk wisdom that sounds quantitative, which is the most persuasive kind, and it survives because everybody can name one example. A single revival proves nothing about a period. You need a long series of measurements before the word cycle means anything at all.

Six dress dimensions, measured year by year

That series exists. In 1940 the University of California Press published Three Centuries of Women's Dress Fashions by Jane Richardson and A. L. Kroeber, as number two in volume five of Anthropological Records. Kroeber was an anthropologist rather than a fashion writer, and he approached dress the way he would have approached any other cultural record, which is to say by measuring it.

The pair worked from fashion plates and portraits of formal evening dress, building year-by-year figures from 1787 to 1936 and thinner data reaching back to 1605. They took six dimensions: skirt length, skirt width, the position of the waist, the width of the waist, the depth of the neckline, and its width. Each was recorded as a proportion of the total height of the figure rather than in inches, so an engraving and an oil painting could be compared without the size of the image distorting anything.

The wave lengths the measurements produced

Their table seventeen sets the six results side by side, and the pattern in it is nothing like a generation.

DimensionMean wave length, years
Skirt length100
Skirt width100
Waist length71
Waist width93
Neckline depth71
Neckline width154

The average of the six comes to ninety-eight years. Richardson and Kroeber summarized it as a proportion taking about half a century to travel from one extreme to the other, long or full at one end and short or narrow at the other, then about another half century to travel back. They were careful about the finding rather than triumphant, saying plainly that they could not tell how much weight the clustering of those wave lengths around a century deserved. The direction of the result was still clear enough: the basic proportions of dress move slowly, which cut against the assumption of their own time that fashion changes at random and at speed.

Laver put twenty years in a different column

The other attempt came three years earlier from James Laver, a curator at the Victoria and Albert Museum, in his 1937 book Taste and Fashion. Laver wrote a ladder of how the same garment reads as time passes. Indecent ten years before its moment, shameless at five, daring at one, smart while current, dowdy a year later, hideous after ten, ridiculous after twenty, amusing after thirty, quaint after fifty, charming after seventy, romantic after a hundred, and beautiful only after a hundred and fifty years.

Read that ladder against the popular rule and the mismatch is hard to miss. Twenty years is where Laver placed the bottom of the curve. It is the point at which a garment looks most absurd, not the point at which it comes home in triumph. Whatever people mean by the twenty-year rule, they are describing the ugliest rung on the only timeline anyone bothered to write down.

Laver's ladder is a personal observation rather than a measured result, and researchers treat it that way. In a 2014 paper for the International Textile and Apparel Association, Natalie Potts and Genna Reeves-DeArmond of Oregon State University describe it as a reflection rather than a theory built on data, and argue it needs revision because 1990s trends were already back in shops in 2013 and 2014, well ahead of anything Laver's schedule allowed.

Surface returns fast, structure does not

Both findings fit together once you separate two things that get called the same word. Surface details move quickly, because they cost a factory almost nothing to change. Structure moves slowly, because it decides what the whole garment is.

Timing a purchase against a slow wave

Turn this into a rule you can use in a shop. When you are buying structure, and coats, tailoring, and anything that sets a waist or a shoulder count as structure, buy toward the middle of the wave rather than either extreme. A mid-rise, a moderate leg width, and a shoulder that follows your own will stay wearable through most of a decade, because a proportion needs about fifty years to walk from one end of its range to the other.

When you are buying surface, price it as a short-life purchase and pick accordingly. A revived color, wash, or toe shape can be genuinely enjoyable and there is no reason to avoid it, provided you paid an amount that a couple of seasons justifies. Then apply the same split to your own wardrobe before you store anything. Keep structural pieces in neutral proportions, because those come back. Let the surface pieces go, because by the time they return the version in shops will be someone else's, and Laver already told you what year twenty will look like.