The Great Shift: When Services Overtook Everything

The American economy has two great halves. One makes physical things — food, fuel, buildings, and manufactured goods. The other provides services — healthcare, finance, software, legal advice, meals, and everything else that cannot be dropped on your foot. For most of the 20th century, these two halves existed in rough proportion. Factories employed millions. Construction boomed. Mining and agriculture anchored entire regions.

Then the balance broke.

Over the 27 years from 1997 to 2024, the goods-producing share of American GDP fell from 22.5% to 16.6% — a decline of 5.9 percentage points. The services-providing share rose from 64.2% to 72.2% — a gain of 8.0 points. The gap between the two widened from 41.7 percentage points to 55.6 points. Services-providing industries now generate 4.3 times as much value added as goods-producing industries.

But the most striking feature of this transformation is not its size. It is its rhythm. The shift did not accumulate steadily, year by year, like sand through an hourglass. It came in three sharp drops — each one triggered by a recession, each one permanent.

The chart tells a story of three acts. In the late 1990s, goods and services moved in near-lockstep — goods drifting slowly lower from 22.5% to 21.7%, services edging up from 64.2% to 65.4%. The economy was growing so fast that both halves expanded in absolute terms, even as their relative shares diverged only slightly.

Then the dot-com bubble burst. In 2001, the goods-producing share plunged from 21.7% to 20.6% in a single year — a 1.1 percentage point drop that erased two years of stability. Manufacturing was hit hardest, losing production as the tech investment cycle collapsed. Services, meanwhile, barely flinched. By 2002, the goods share fell further to 19.9%. The subsequent recovery brought goods back only to 20.9% by 2006 — still 1.6 points below where they had been before the recession struck.

The pattern would repeat, twice more, each time more brutally.

The table below captures what may be the most important structural dynamic in the American economy over the past three decades. In each recession, goods-producing industries lost a significant chunk of their GDP share. In each subsequent expansion, they recovered only a fraction of what they had lost. The rest was permanently transferred to services.

*The 2020 pandemic recovery was the only case where goods fully recaptured their lost share — goods briefly hit 17.4% in 2022 as supply-chain disruptions boosted domestic production and manufacturing pricing power. By 2024, however, goods had fallen back to 16.6%, well below the pre-pandemic 17.1%.

The 2001 recession was the first ratchet. Goods lost 1.9 percentage points of GDP share and recovered only 0.3 points in the subsequent expansion. The financial crisis was far worse: goods plunged from 20.9% to 18.4% — a 2.5-point drop — and clawed back just 0.4 points before the next downturn. By the time the pandemic arrived, goods had already been so diminished that the 0.8-point drop was smaller in absolute terms, though the pandemic’s brevity and the supply-chain pricing surge allowed an unusually complete (but temporary) rebound.

The cumulative result: 5.9 percentage points of GDP share permanently transferred from goods to services over 27 years. That is roughly $1.7 trillion per year in value added that now flows through service industries instead of factories, mines, and construction sites.

The shift from goods to services was not an abstract reclassification. It played out through specific industries gaining and losing ground. The chart below shows the net change in GDP share for the major sectors from 1997 to 2024. One sector dominates the losses. Several services split the gains.

Manufacturing lost 6.3 percentage points of GDP share — falling from 16.1% in 1997 to 9.8% in 2024. This is the single largest shift in the entire BEA dataset. To put it in perspective, manufacturing’s loss is larger than the total GDP share of the information sector (5.4%) or healthcare (7.5%). An entire major industry’s worth of economic share migrated elsewhere.

On the gaining side, professional and technical services added 2.2 percentage points — the largest single winner. This is the sector of consulting firms, software developers, accountants, architects, and engineers. It has risen every single year since 1997 without a single interruption, a record no other sector can match. Real estate gained 1.7 points. Healthcare added 1.5 points. Finance gained 0.9 points. Information, despite the tech revolution, added just 0.8 points — a puzzle we will explore in Episode 5.

Agriculture lost 0.4 points, falling from 1.3% to 0.9% of GDP. Construction gained 0.5 points. Mining remained roughly flat when measured over the full period, though it swung wildly in between — surging during the 2010s shale boom and collapsing when oil prices crashed.

It would be easy to look at these numbers and conclude that American manufacturing is in terminal decline. That would be wrong.

Manufacturing real output grew 71% from 1997 to 2024. American factories produce far more than they did a generation ago — more cars, more chemicals, more aerospace components, more semiconductor equipment. The sector’s share of GDP fell not because it shrank, but because services grew faster. Much faster. Professional services roughly tripled in real terms. Healthcare doubled. Finance expanded relentlessly.

This distinction matters enormously for policy. A sector that is shrinking in absolute terms needs rescue. A sector that is growing but losing relative ground needs a different conversation entirely — one about productivity, wages, and comparative advantage rather than decline and decay.

The services economy did not kill manufacturing. It outran it. Services proved more responsive to the forces that defined the modern economy: rising incomes, an aging population, the explosion of digital technology, and the growing complexity of business itself. Every dollar of GDP growth increasingly flowed toward industries that help, heal, advise, finance, and entertain rather than industries that dig, build, and assemble.

The recession ratchet is not an accident. There are structural reasons why downturns permanently reshuffle the economy toward services.

Goods demand is cyclical; services demand is sticky. When a recession hits, consumers postpone buying cars, appliances, and houses. Businesses cancel factory orders and halt construction projects. But people still go to the doctor, still pay rent, still need legal counsel, still use their phones. The demand shock is asymmetric — it hits goods hard and services lightly.

Recovery favors efficiency over headcount. When goods-producing firms rebuild after a recession, they invest in automation and lean production rather than rehiring to pre-recession levels. Each recovery produces the same or greater output with fewer workers and less share of the total economy. Services firms, by contrast, expand by adding people — more consultants, more nurses, more software developers — which translates directly into higher value added.

Global competition is fiercer for goods. Each recession triggers a round of offshoring as manufacturers seek cheaper production. The 2001 recession accelerated outsourcing to China. The 2008 crisis pushed more supply chains overseas. These shifts are rarely reversed when the economy recovers.

The result is a one-way gate. Goods lose share in the downturn and do not fully recover it in the expansion. Over three recessions, this ratchet transferred nearly 6 percentage points of GDP from goods to services — a structural transformation disguised as a business cycle.

The 27-year arc described in this episode raises an obvious question: which industries within these two great halves are driving the shift? The goods-producing decline is dominated by one sector — manufacturing — which accounts for 6.3 of the 5.9 net percentage points lost (offset by construction’s 0.5-point gain). But within manufacturing, the story is more complicated. Some sub-industries collapsed. Others thrived. The overall decline masks a radical internal reshuffling that deserves its own episode.

In Episode 3, we zoom into manufacturing’s long retreat — the sectors that vanished, the sectors that held, and the surprising pockets of growth inside America’s shrinking industrial core.