Manufacturing’s Long Retreat

In 2020, for the first time in the history of the Bureau of Economic Analysis data, manufacturing fell below 10% of American GDP. It was not a one-year anomaly driven by pandemic shutdowns. Manufacturing hit 10.1% in 2019, dipped to 9.8% during the Covid recession, bounced back to 10.4% amid the 2022 inflation surge — and then fell right back to 9.8% in 2024. The threshold has been breached, and it does not appear to be coming back.

The decline from 16.1% in 1997 to 9.8% in 2024 represents a loss of 6.3 percentage points of GDP share — the single largest shift of any industry in the BEA data. No other sector gained or lost that much ground. It is the defining structural story of the American economy over the past three decades.

And yet the story is not what most people assume. Manufacturing is not dying. It is not even shrinking. In real (inflation-adjusted) terms, manufacturing value added rose from $1,379 billion in 1997 to $2,361 billion in 2024 — a gain of 71.2%. Factories are producing far more than they did a generation ago. They are simply being outrun by an economy that grew even faster around them.

The chart tells the story in three acts. Act one: the cliff. Between 1997 and 2002, manufacturing’s share plunged from 16.1% to 12.0% — a loss of 4.1 percentage points in just five years. The dot-com bust exposed how much of the late-1990s manufacturing boom was tied to the technology buildout. When the buildout stalled, factories went dark.

Act two: the plateau. From 2003 to 2007, manufacturing held steady at roughly 12% of GDP. The housing boom and consumer spending kept factories humming. But the financial crisis delivered a second blow, knocking the share down to 10.6% by 2009. Manufacturing recovered to about 10.9% during the 2010s expansion — but never regained its pre-crisis level.

Act three: the breach. The pandemic pushed manufacturing below 10% for the first time in 2020. The 2021–2022 inflation surge briefly lifted the nominal share back above 10% (higher prices temporarily inflated manufacturing’s dollar contribution), but as inflation moderated, the share fell back to 9.8% in 2024. The structural trend has reasserted itself.

Manufacturing is not one industry. It is seventeen sub-industries tracked by the BEA, ranging from food processing to computer chip fabrication. The aggregate decline from 16.1% to 9.8% of GDP masks enormous variation within: some sub-industries were nearly wiped off the map, while others held their ground or even grew.

Computer and electronic products is the poster child of the paradox. Its share of GDP fell from 2.3% in 1997 to just 1.0% in 2024 — a decline of more than half. But in real terms, output exploded from $22.1 billion to $287.2 billion, a gain of roughly 1,200%. This is the productivity revolution distilled into a single data point: American factories are producing exponentially more computing power, at exponentially lower prices, generating enormous consumer surplus that the GDP statistics barely capture. The share fell because prices collapsed even as output soared.

Chemical products — pharmaceuticals, petrochemicals, plastics feedstocks — is the one that held. Its share barely moved, from 2.0% to 1.9% over 27 years. In real terms, output grew from $298 billion to $476 billion. Chemicals is the steady anchor of American manufacturing: always essential, never glamorous, quietly growing at roughly the pace of the overall economy.

Motor vehicles tells a grimmer story. The iconic American industry saw its GDP share cut in half, from 1.3% to 0.6%. The 2008–2009 auto crisis — the bankruptcy of General Motors and Chrysler, the collapse of the Detroit supplier network — permanently reduced the industry’s footprint. The recovery was real but incomplete. And the ongoing shift to electric vehicles is reshaping the supply chain yet again.

The table reveals a clear pattern. The survivors — chemicals, petroleum, food — are industries tied to physical necessities. People need drugs, fuel, and food regardless of what the tech sector is doing. These sub-industries held their share because demand for their products grows with the population and because their prices did not collapse the way electronics prices did.

The biggest losers fall into two categories. First, the productivity victims: computer and electronics manufacturing lost 1.3 percentage points of GDP share despite its output exploding, because its prices fell so dramatically that the nominal value of its contribution shrank relative to the economy. Second, the competition victims: textiles, furniture, printing, and paper lost share because their production moved overseas or because demand for their products simply declined. Nobody prints as much as they used to. Nobody buys American-made furniture the way they once did.

Petroleum and coal is the lone gainer — rising from 0.6% to 0.7% — a testament to the shale revolution that turned the United States into the world’s largest oil producer after 2018.

The central paradox of American manufacturing is this: the sector is producing more than ever, employing fewer people than ever, and shrinking as a share of the economy. All three statements are simultaneously true, and none of them contradicts the others.

Real manufacturing output grew 71.2% from 1997 to 2024. But total real GDP grew even faster — from roughly $12.6 trillion to $23.3 trillion in chained 2017 dollars, a gain of about 85%. Manufacturing grew handsomely in absolute terms but fell behind the overall pace. The services economy — healthcare, professional services, finance, information — simply expanded faster.

Meanwhile, manufacturing employment fell from 17.4 million workers in 1997 to roughly 12.9 million in 2024. The sector produces 71% more output with 26% fewer workers. That is a productivity revolution — but it is also a human story of displaced workers, hollowed-out factory towns, and a political backlash that reshaped American elections.

The computer and electronics sub-industry captures this paradox in its purest form. Real output rose 1,200%. GDP share fell by more than half. The industry is simultaneously the greatest productivity success story in American manufacturing history and the largest single contributor to manufacturing’s declining share of GDP. It produces staggering quantities of value at staggering efficiency — and the statistics reward it by counting it as a smaller and smaller piece of the pie.

Manufacturing’s story is the anchor of this series because it is the story that defines the transition. Every percentage point that manufacturing lost had to go somewhere. Some of it went to professional services, some to healthcare, some to finance and real estate. The next several episodes will trace those destinations.

But first, in Episode 4, we go deeper into the wreckage. Manufacturing’s aggregate decline of 6.3 points is dramatic — but it is not the only sector that lost ground. Several entire industries shrank so much they effectively vanished from the GDP map. Episode 4 tells the story of the industries that disappeared.