The Staircase That Only Goes Down

Economists have a term for systems that lose capacity after each shock and never fully recover: hysteresis. It’s borrowed from physics, where certain materials retain the deformation caused by stress even after the stress is removed. Bend a paperclip back and forth, and each cycle leaves it a little more warped, a little closer to snapping. The American labor market has been doing something strikingly similar for a quarter century.

In 2000, at the peak of the dot-com boom, the gross job creation rate stood at 8.3% — meaning 8.3% of all private-sector employment was in positions that hadn’t existed the previous quarter. Firms were expanding, new businesses were opening, and the economy was churning through workers at a pace that, in retrospect, would prove to be the modern high-water mark. Then came the 2001 recession. The gain rate fell to 7.1% by 2003. When the economy recovered and hiring boomed during the mid-2000s housing bubble, the gain rate climbed back to… 7.0%. Not 8.3%. Not even close. The recovery restored GDP, restored employment, restored the stock market. It did not restore dynamism.

Then came the Great Recession. The gain rate collapsed to 5.4% in 2009 — the lowest on record. Through the longest expansion in American history, from 2010 to 2019, it clawed its way back to 5.9%. Then Covid hit, the rate fell again to 5.5%, a historic surge of new business formation in 2021-2022 briefly pushed it to 6.9%, and by 2025 it had settled back to 5.6%. Each recession carved a new, lower floor. Each recovery found a new, lower ceiling. The staircase only goes down.

The loss rate tells the same story in mirror image. Firms were destroying 7.4% of employment per quarter in 2000. By 2019, that rate had fallen to 5.6%, and by 2025, to 5.4%. This isn’t a story about destruction outpacing creation — it’s a story about both sides of the engine slowing simultaneously. Less creation and less destruction. Less churn. Less reallocation. Less of the economic metabolism that, as we saw in Episode 1, is the essential mechanism of growth.

The gross job creation rate is actually two rates welded together: expansions and openings. Expansions measure the hiring at existing establishments that added workers. Openings measure the jobs created by brand-new establishments in their first quarter of existence. Together, they constitute every job gained in the economy. Separating them reveals something unexpected about where the decline in dynamism actually lives.

The expansion employment rate — jobs gained at existing firms that grew — fell from 6.5% of total employment in 2000 to 4.4% in 2025. That’s a decline of 2.1 percentage points, or roughly a third of its starting value. Meanwhile, the opening employment rate fell from 1.8% to 1.2%, a decline of 0.6 percentage points. Both fell, but expansions drove three-quarters of the total decline. The opening rate dropped, but from a much smaller base, and by 2024 it had actually stabilized and even ticked up slightly to 1.2% from its 2020 trough of 1.0%.

This distinction matters because the public conversation about declining dynamism tends to focus almost entirely on startups. The narrative is familiar: big companies are crushing small ones, barriers to entry are rising, the entrepreneurial spirit is dying. There’s truth in that story, as we’ll see in Episode 3. But the data says the bigger problem isn’t that fewer businesses are starting — it’s that existing businesses are growing less. The expansion rate’s collapse from 6.5% to 4.4% means that in a workforce of roughly 130 million, about 2.7 million fewer jobs per quarter are being created by existing firms growing their payrolls than would have been created at the 2000 rate. That’s the equivalent of losing a medium-sized city’s entire employment base every three months — not from layoffs, but from hiring that simply never happened.

On the destruction side, the same pattern holds. The contraction rate (existing firms shrinking) fell from 5.8% to 4.4%, while the closing rate (establishments shutting down entirely) held steadier, dropping from 1.6% to 1.0%. Firms are both growing less and shrinking less. They’re sitting still. In a healthy economy, you want businesses constantly experimenting — hiring a few people to test a new product line, trimming a department that isn’t working, scaling up what succeeds. The data shows less of all of that.

If the decline in dynamism is primarily about existing firms growing less, an obvious question follows: which firms? The BDS breaks establishments into nine size classes, from businesses with 1-4 employees up to those with 1,000 or more. The pattern across these categories is striking and, in some ways, counterintuitive.

The smallest firms — those with 1 to 4 employees — experienced the largest absolute decline in their job creation rate: from 18.9% in 2000 to 14.8% in 2024, a drop of 4.1 percentage points. These are the corner shops, the solo consultancies, the three-person landscaping crews. Their high creation rate reflects the fact that when a firm with two employees hires a third, that’s a 50% expansion, which registers as enormous dynamism in percentage terms. But the rate has been steadily compressing. The next size classes followed suit: firms with 5-9 employees dropped 2.6 points, those with 10-19 dropped 2.5. The pattern is monotonic — the smaller the firm, the bigger the decline.

At the other end of the spectrum, firms with 1,000 or more employees saw their gain rate fall from 3.6% to 2.3%, a decline of only 1.3 percentage points. These are the Amazons, the hospital chains, the defense contractors. They were never particularly dynamic to begin with — their sheer size makes large percentage moves mathematically difficult — and they’ve stayed roughly where they were. The gap between the smallest and largest firms’ creation rates narrowed from 15.3 percentage points in 2000 to 12.5 points in 2024. The economy is converging toward the metabolism of its biggest players.

Why would small firms slow down more? Several forces are likely at work. Regulatory burden tends to fall disproportionately on small firms, which lack compliance departments and legal teams. Health insurance costs, which more than tripled in real terms between 2000 and 2024, hit small employers hardest because they can’t negotiate the volume discounts that Fortune 500 companies command. The rise of e-commerce and platform economies created winner-take-all dynamics that made it harder for local businesses to grow into regional ones. And the consolidation of banking — the number of FDIC-insured banks fell from 8,315 in 2000 to about 4,500 by 2024 — reduced the availability of the kind of relationship lending that small firms depend on. Whatever the cause, the smallest firms are disproportionately responsible for the economy’s declining dynamism.

The decline in dynamism is not confined to a few struggling sectors. Every one of the 13 major industry groups in the BDS saw its job creation rate fall between 2000 and 2024. Not a single one bucked the trend. But the severity varies enormously, and the pattern reveals something about the structural forces at work.

The biggest losers, in percentage terms, were wholesale trade (−40%), transportation and warehousing (−39%), and information (−39%). These are precisely the industries most affected by the twin revolutions of e-commerce and digital automation. Wholesale trade — the middlemen who warehoused and distributed physical goods — saw their gain rate plummet from 7.2% to 4.3%. Amazon, Shopify, and direct-to-consumer brands didn’t just displace individual wholesalers; they reduced the industry’s need to churn through establishments at all. Fewer new distribution firms open when a single fulfillment network can serve the entire country. Information (which includes publishing, broadcasting, telecom, and data processing) saw its gain rate fall from 8.5% to 5.2% as the internet consolidated media and communications into a few dominant platforms.

At the other end of the spectrum, education and health services saw the smallest decline: from 5.7% to 5.0%, a relative drop of just 12%. This makes intuitive sense. Hospitals don’t get disrupted by apps. Schools don’t get automated away. The demand for nurses, therapists, and teachers is driven by demographics — aging, birth rates, chronic disease prevalence — not by market competition. Similarly, leisure and hospitality (−21%) proved relatively resilient because restaurants and hotels are inherently local, physical, and difficult to consolidate. You can’t eat a meal through a smartphone.

Natural resources and mining present an interesting case: the sector’s gain rate plummeted from 20.5% to 12.7%, a fall of 7.8 percentage points — the largest absolute decline of any sector. But this reflects the unique volatility of commodity extraction. In 2000, the oil patch was in a hiring frenzy. By 2024, the shale revolution had matured, rig counts had stabilized, and the industry had learned to produce more barrels with fewer workers. The sector’s dynamism wasn’t destroyed by sclerosis; it was disciplined by efficiency. The distinction matters, and we’ll return to it in Episode 6.

Every state in the union saw its job creation rate decline between 2000 and 2024. Zero exceptions. But the magnitude of decline varies by a factor of more than three, and the geographic pattern tells a story about cost, regulation, and the changing structure of American capitalism.

The states that lost the most dynamism cluster along two corridors. The first is the high-cost Northeast: New Jersey leads the nation with a 46.6% decline in its gain rate (from 10.3% to 5.5%), followed by Massachusetts (−31%), Pennsylvania (−31%), and New York (−33%). The second is the West Coast: Washington (−44%), California (−36%), and Colorado (−32%). These are states where housing costs, commercial rents, and regulatory complexity have risen the most over the past quarter century. When it costs $5,000 a month to rent a storefront and six months to get a building permit, fewer entrepreneurs take the plunge, and existing businesses think twice before adding a second location.

The states that retained the most dynamism are a mix of low-cost Southern and rural states: West Virginia (−13%), Idaho (−14%), Mississippi (−14%), South Carolina (−15%), and Alabama (−15%). These aren’t exactly economic powerhouses — in fact, several of them rank near the bottom in absolute job creation rates. But they started from a lower base of dynamism and had less to lose. Idaho is the notable exception: it maintained a high absolute gain rate (7.9% in 2024, among the nation’s highest) while experiencing a relatively modest decline, reflecting the state’s genuine boom in tech migration and new business formation during the 2020s.

Florida illustrates the complexity of this story. It saw a steep 38% decline in its gain rate (from 9.4% to 5.8%), placing it among the biggest losers. Yet Florida has been the destination of choice for business relocations, the epicenter of pandemic-era migration, and a state that prides itself on low taxes and light regulation. How can a booming state be losing dynamism? Because dynamism isn’t the same as growth. Florida is growing, but it’s growing more efficiently — with fewer new establishments per unit of job growth, larger average firm sizes, and more incumbent expansion relative to startup churn. The state is getting bigger while getting less turbulent, which is exactly the national pattern writ large.

Rates are one thing. But what about the actual number of jobs being created? The private sector employed roughly 108 million workers in 2000 and about 133 million in 2025. With a bigger workforce as the denominator, even a declining rate might still produce more absolute job flows. It doesn’t.

In the first quarter of 2000, 9.0 million jobs were created in a single quarter. In Q1 2025, just 7.4 million were. The economy is 23% bigger but creating 17% fewer jobs per quarter. That’s the absolute paradox of declining dynamism: the economy grows, employment rises, GDP expands, yet the raw machinery of job creation is producing less output than it did a quarter century ago. It’s as if you upgraded from a four-cylinder engine to a six-cylinder one but somehow lost horsepower in the process.

The loss side tells a similar story. In Q1 2000, 8.1 million jobs were destroyed. By Q1 2025, that had fallen to 7.2 million. This isn’t entirely bad news — fewer destroyed jobs means fewer workers displaced, fewer families disrupted, less economic pain in the short term. But it also means fewer resources being freed from low-productivity uses to flow toward higher-productivity ones. When a mediocre business stays open because the competitive pressure that would have killed it in 2000 no longer exists, that’s capital, labor, and rent locked into a suboptimal use. Economists call this “zombie” behavior, and declining destruction rates are one of its signatures.

The post-pandemic period offers the most dramatic illustration of the paradox. In 2021 and 2022, gross job gains surged to 8.3 million and 8.5 million per quarter, respectively — levels not seen since the early 2000s. For a moment, it looked like the pandemic might have permanently reset the economy’s dynamism. But by 2024, gains had fallen back to 7.7 million, and by Q1 2025, to 7.4 million. The pandemic surge was a one-time reconfiguration, not a permanent revival. The staircase resumed its descent.

Declining dynamism sounds abstract, but its consequences are concrete. When fewer businesses expand, fewer workers get poached by growing firms, and that removes one of the most powerful mechanisms for wage growth. An employee at a stagnant firm can ask for a raise, but an employee at a stagnant firm who has an offer from a growing competitor will get a raise. The decline in dynamism is correlated with the well-documented stagnation of real wages for middle-income workers between 2000 and 2019, and it is not a coincidence.

Less dynamism also means less geographic mobility. When firms in boomtown industries are hiring aggressively, workers move to where the jobs are. When the churn rate falls, there are fewer signals pulling workers from declining regions to thriving ones. The rate of interstate migration in the United States fell from about 2.7% per year in the early 2000s to about 1.5% by 2023 — a decline that tracks the dynamism data almost exactly. The labor market is not just churning less; it is sorting less efficiently.

For investors, declining dynamism has a specific implication: incumbency advantages are widening. When the rate of business openings falls and the rate of competitive turnover drops, the firms that already dominate their markets face less pressure from below. This is one reason why industry concentration has increased across most of the economy, why profit margins for large firms have expanded even as GDP growth has slowed, and why the stocks of dominant companies have outperformed the broad market so consistently in the 2010s and 2020s. The creative destruction machine is still running, but the “destruction” part is running slower than the “creative” part, and that favors the established over the insurgent.

The most unsettling implication is for the future. If each recession permanently ratchets dynamism lower, and if we are due for another recession at some point (as we inevitably are), then the next downturn will start from an already low base. The post-pandemic surge showed that extreme shocks can temporarily revive dynamism, but the effect was fleeting. Unless something structural changes — regulatory reform, a new wave of enabling technology, a dramatic reduction in the cost of doing business — the staircase will keep descending. The question is not whether American capitalism is less dynamic than it was a generation ago. It is. The question is whether the decline is a temporary phase or a permanent condition.