The Industries That Vanished

In Episode 3, we told the story of manufacturing’s long retreat — how a sector that once defined American economic identity fell from 16.1% of GDP in 1997 to 9.8% in 2024. But that headline obscured a crucial distinction. Most of that retreat was relative. Manufacturing’s real output actually grew 71% over those 27 years. It simply could not keep pace with services, finance, and real estate, which grew faster.

Most manufacturing sub-industries followed that pattern. Chemicals grew 93% in real terms. Computer and electronic products grew over 1,200%. Food and beverage manufacturing grew 32%. These sectors added real value to the economy — they just got outrun.

But some industries didn’t just lose their share of a growing pie. The pie grew while their slice physically shrank.

Textiles, paper, printing, and furniture all produce less real output today than they did in 1997. Not less as a share of GDP — less in absolute terms, measured in constant 2017 dollars. These are the industries that truly vanished: trade-exposed, labor-intensive sectors that could not automate fast enough to survive globalization, and in some cases, were made obsolete by technology itself.

The chart tells a story of four industries on the same downward trajectory, but each with its own rhythm. Textiles began declining almost immediately, falling steadily from $26.3 billion in 1997 to $13.4 billion in 2024 — a loss of 49% of its real output. There was no plateau, no stabilization. The North American Free Trade Agreement had opened the door to Mexican production in 1994, and China’s entry into the WTO in 2001 blew it wide open. American textile mills simply could not compete with labor costs a fraction of their own.

Paper followed a slower, steadier decline — from $82.2 billion in 1997 to $56.8 billion in 2024, a loss of 31%. This was not primarily about trade. It was about email. The paperless office, long predicted and long mocked, actually happened — just gradually enough that nobody declared victory. Office paper demand peaked around 2000 and has fallen every year since. Newsprint consumption collapsed with the newspaper industry. Packaging paper survived, but it could not offset the losses.

Printing is the most dramatic story. It actually grew through the early 2000s, reaching $44.3 billion in 2007. Then the internet ate it. Newspapers, magazines, catalogs, direct mail — the entire ecosystem of printed media contracted in the space of a decade. By 2024, printing output had fallen to $28.7 billion, down 20% from 1997 and 35% from its peak. The printing industry did not lose a trade war. It lost a technology war.

Furniture held roughly steady through 2007 at $38.3 billion, then collapsed. The financial crisis gutted housing starts, and furniture demand went with them. But unlike housing, furniture never fully recovered — in part because Chinese imports filled the gap. By 2024, real output had fallen to $23.6 billion, down 38% from 1997.

The table reveals a spectrum of decline. At one end, textiles and apparel lost roughly half their real output — the most severe contractions of any manufacturing sub-industry. At the other end, plastics and rubber barely survived, eking out a 5% gain over 27 years in an economy that grew 127%. And then there is the outlier.

Primary metals — steelmakers, aluminum smelters, foundries — doubled their real output. This was the sub-industry everyone expected to die. Steel towns became a byword for deindustrialization. But the BEA data tells a different story: primary metals produced $40.8 billion in real value added in 1997 and $92.6 billion in 2024. The sector consolidated, automated, and pivoted toward specialty and high-strength alloys that Asian mills could not easily replicate. It is a reminder that trade exposure alone did not determine which industries survived. What mattered was whether the survivors could move upmarket faster than imports could follow.

But even primary metals lost GDP share — from 0.6% to 0.3%. Growing 127% in an economy that also grew 127% means you merely kept pace. And every other sub-industry on this list failed to do even that.

The bar chart puts the vanishing industries in their full context. Computer and electronic products grew over 1,200% in real terms — a transformation so extreme it distorts any chart that includes it. Primary metals grew 127%. At the other end, apparel and textiles each lost roughly half their real output.

The pattern is clear. The industries that survived and thrived were capital-intensive, technology-driven, and produced goods that could not easily be shipped from a low-wage factory overseas. Semiconductors require billion-dollar fabrication plants. Specialty steel requires metallurgical expertise built over decades. Pharmaceuticals require FDA approval processes that take years.

The industries that vanished shared opposite traits. They were labor-intensive. Their products were commoditized. They competed primarily on cost. And once trade barriers fell, cost was a competition they could never win.

Each vanishing industry has its own proximate cause of death, but they share a common underlying pathology: exposure to low-cost competition without a defensible technological moat.

Textiles and apparel were the first to fall because clothing is the textbook case of a commodity product. A T-shirt sewn in Bangladesh is functionally identical to one sewn in North Carolina. When China joined the WTO in 2001 and the Multi-Fiber Arrangement expired in 2005, the last quotas protecting American textile jobs disappeared. U.S. textile employment fell from 630,000 in 1997 to under 230,000 by 2024. The mills that survived specialize in technical fabrics — materials for the military, automotive, and medical industries — where specifications matter more than price.

Paper was killed by a revolution in how people communicate. Business correspondence went electronic. Newspapers lost readers, then advertisers, then pages. Magazine circulations collapsed. The paper industry’s saving grace is packaging — the cardboard boxes that carry everything ordered online — but packaging alone cannot offset the loss of all the paper that once carried information.

Printing died with its customers. When newspapers shrank from 60 pages to 20, the printing presses ran fewer hours. When catalogs moved online, the direct-mail industry contracted. When magazines folded, the high-end printing shops that produced them closed. Printing’s output peaked in 2007 — the year the iPhone launched and the last year before the smartphone gutted print media.

Furniture was a hybrid casualty — partly trade, partly the housing crisis. Chinese furniture exports to the U.S. surged in the 2000s, wiping out much of the domestic industry in North Carolina and Virginia. Then the 2008 housing collapse destroyed demand. And when demand recovered, it was increasingly met by imports and by flat-pack retailers like IKEA, whose business model bypasses domestic manufacturing entirely.

These are not just lines on a chart. They are maps of regional devastation. The textile and apparel decline was concentrated in the American Southeast — the Carolinas, Georgia, Alabama — where entire towns had been built around a single mill. When the mill closed, the town’s largest employer, its tax base, and its reason for existing all vanished together.

Furniture manufacturing was centered in western North Carolina and parts of Virginia. The High Point furniture market remains the industry’s annual showcase, but much of the furniture on display is now manufactured overseas. Paper and printing losses were more dispersed, but they hit hardest in the Upper Midwest and Northeast — places like Appleton, Wisconsin, and the mill towns of Maine and upstate New York.

The human cost is measured not just in lost output but in lost options. A textile worker in Gastonia, North Carolina, in 1997 had a clear career path: start at the mill, learn the trade, work for thirty years, retire with a pension. By 2010, the mill was closed, the pension was frozen, and the available jobs paid half as much. The aggregate GDP data cannot capture this, but any honest accounting of the “new economy” must acknowledge it.