In December 1993, a men’s dress shirt sold at a US department store might run thirty dollars. Run the Bureau of Labor Statistics’ apparel price index for that same month, CUUR0000SAA, and it reads 132.6, on a scale where 1982 to 1984 averages 100. In December 2025, the same index reads 128.1. Lower. Meanwhile the government’s broader measure of everything Americans buy, the all-items CPI, climbed from 145.8 to 324.1 over that span, more than doubling. Rent, food, and gasoline all cost roughly twice as much. A shirt costs, in nominal dollars, slightly less than it did when Bill Clinton was a year into his first term.
That gap is two named government series pulling apart for three decades straight, the first hard number behind a complaint that usually gets dismissed as nostalgia.
What “got worse” can actually mean
“Clothes got worse” is vague enough to argue about forever, so the useful move is to break it into things a data series can settle. Three claims get tested here: the material changed, the price fell in real terms, and more of it gets made per person on earth. Each has a named source behind it. Where a fourth claim, that stitch density or seam strength declined, would need a series nobody publishes at this resolution, it gets left out rather than asserted on vibes.
The three that survive are not independent observations that happen to point the same way. They describe one supply chain responding to one set of incentives over the same thirty years: cheaper input fiber, a price ceiling that would not move, and a market that kept ordering more units to hit it. Reading them separately risks treating each as a coincidence. Read together, they are closer to a single mechanism with three readouts.
The fiber mix moved to plastic
Textile Exchange’s Materials Market Report 2024, covering 2023 production, puts polyester at 57 percent of global fiber output on its own. Add the other virgin fossil-based synthetics (nylon, acrylic, and the rest) and the fossil-derived share reaches about 60.5 percent of everything spun into thread that year. Cotton, the fiber that dominated clothing for most of the twentieth century, comes in around 19.7 percent. Recycled polyester accounts for 12.5 percent of the total, which means it is counted inside that 57 percent, not on top of it: most polyester, recycled or not, is still polyester, a thermoplastic spun from petroleum.
Polyester is not automatically inferior. It resists wrinkling, dries fast, and holds color well, which is why performance wear leans on it deliberately. But a majority- polyester wardrobe behaves differently on a body than a majority-cotton one did: less breathable, more prone to pilling under friction, and it does not biodegrade on any timeline that matters to a landfill. The shift from a natural-fiber-majority supply chain to a plastic-majority one is a real material change, not a matter of brand quality slipping. The whole input base moved.
The price fell because something else moved with it
Apparel is one of the only major US consumer categories where the sticker price has sat still while general inflation kept climbing for three decades, and the retailer is not being generous. A garment that costs the same in 2025 dollars as it did in 1993, while labor, cotton handling, and shipping all got more expensive in nominal terms, was paid for somewhere. Textile Exchange’s fiber-mix numbers point at part of the answer: polyester is cheaper to produce at scale than combed cotton, and a supply chain that shifted toward it bought itself room under the price ceiling. The other part is construction: fewer stitches per inch, thinner cuts of fabric, and manufacturing that moved to wherever labor cost the least, repeatedly, over thirty years. The BLS series does not itself measure any of that directly. It measures what shoppers paid, and what shoppers paid, adjusted for inflation, fell hard.
More clothing gets made, and more of it per person
The volume side confirms the pattern rather than explaining it away. Textile Exchange counted 116 million tonnes of fiber entering global production in 2022 and 124 million tonnes in 2023, a single year’s jump of roughly 7 percent. Zoom out further and the Ellen MacArthur Foundation’s 2017 report, A New Textiles Economy, found that clothing production had roughly doubled in the fifteen years leading up to that publication. World population did not double over a comparable stretch. UN figures put it at about 6.13 billion in 2000 and 7.35 billion in 2015, a rise of roughly 20 percent. Production that doubled against a population that grew a fifth means more garments got made per person, on top of the larger total. That is the volume leg of the argument, built from two separately sourced numbers rather than one series that already does the division, so treat the per-capita framing as arithmetic on real data rather than a fourth primary source.
What the data does not show
None of these three series measures durability directly. No government agency publishes a “average wears before failure” index, and no industry report tracks seam strength across decades the way BLS tracks price. The plan behind this piece wanted a fourth leg, a direct durability series, and it does not exist in a form that would survive a source check, so it is left out rather than invented. What the fiber-share, price, and volume data support together is narrower and still real: the material base shifted toward plastic, the price fell in a way that only makes sense if something in the making of the garment also got cheaper, and more units get produced against a population that is not growing nearly as fast. Whether any single shirt in a closet got worse is a question about that shirt. Whether the system that makes shirts changed shape is a question these three agencies and one foundation already answered.
DRESS tees run 6.1 ounces of heavyweight cotton, printed to order rather than overproduced against a forecast, which is one way to sit outside the trend this data describes rather than inside it.
A dress shirt from December 1993 and one bought last week can carry the same number on a price tag adjusted for inflation. They are not made from the same world.
