The Great Rebalancing: Navigating the Paradox of Declining U.S. Business Dynamism and Surging Innovation

Introduction: The Paradox of Stagnant Dynamism and Explosive Innovation

For decades, economists have tracked a handful of metrics that together paint a picture of a nation’s economic vitality: the rate of new business formation, the churn of jobs between firms, and the competitive turnover that rewards the efficient and punishes the laggards. Collectively, these are the hallmarks of “business dynamism,” and by every major measure, the United States has been losing it. Since the 1980s, the share of firms that are startups has fallen from roughly 15 percent to around 8 percent. The job reallocation rate—the sum of all jobs created and destroyed as a share of total employment—has dropped by nearly a third. Industry concentration has risen, with the top 10 percent of firms now commanding a larger share of revenue, profits, and R&D spending than at any point in modern history.

Yet, at the same time, the United States is witnessing an explosion of breakthrough innovations. Artificial intelligence has moved from academic curiosity to a force reshaping every industry. Biotechnology has delivered CRISPR-based gene therapies, rapid mRNA vaccine platforms, and a pipeline of novel drugs. Clean energy technologies—solar, battery storage, advanced nuclear—are scaling faster than almost any forecaster predicted. The paradox is stark: on paper, the economy looks sclerotic; in practice, it is producing transformative advances at warp speed.

The thesis of this article is that the decline in traditional dynamism is real but masks a structural shift in *how* and *where* innovation occurs. The locus of innovation has moved away from broad-based entrepreneurship and toward deep-tech sectors dominated by incumbent firms and highly specialized startups that often get absorbed before they can disrupt. Market concentration, often blamed for stifling competition, operates as a double-edged sword: it suppresses new entry in commoditized industries while channeling huge resources into risky, capital-intensive innovation. Policy, meanwhile, is emerging as a critical rebalancing tool—one that could either accelerate or undermine the next decade of American competitiveness.

[IMAGE: Time-series chart of U.S. startup formation rate (1980–2025) overlaid with key innovation milestones (e.g., launch of ChatGPT, CRISPR breakthrough).]

The Long Decline: Data on Startup Formation, Job Churn, and Concentration

The numbers are unequivocal. According to the U.S. Census Bureau, the startup share of all firms—defined as firms less than one year old—peaked at around 15 percent in the early 1980s and has since drifted downward to roughly 8 percent in the 2020s. This is not a cyclical blip; it is a structural trend that persists through booms and busts. The Brookings Institution has documented a parallel decline in job reallocation, which fell from about 22 percent of total employment in the 1990s to around 15 percent in the wake of the Great Recession and has never recovered.

Why does this matter? Because startups are disproportionately responsible for net job creation and for the kind of creative destruction that keeps productivity growing. When new firms stop forming at historic rates, the economy loses a key mechanism for reallocating resources to their most productive uses.

At the same time, industry concentration has intensified. A seminal 2020 paper by David Autor and colleagues showed that the share of sales and employment accounted for by the top four firms in many sectors has increased markedly, especially in retail, finance, and information technology. The pattern is most extreme in digital markets, where network effects, data advantages, and platform economics create winner-take-most dynamics. These conditions make it harder for “high-impact” startups—those that grow rapidly and eventually go public—to emerge. A 2023 study by Ryan Decker and John Haltiwanger found that the share of high-growth young firms has fallen, and those that do achieve scale are increasingly likely to be acquired by incumbents before they can mount a serious competitive challenge.

Regionally, the story is one of deepening divergence. Startup activity is heavily concentrated in a small number of coastal metros—San Francisco Bay Area, New York, Boston, and to a lesser extent, Seattle and Los Angeles. In the Rust Belt, the South, and rural America, business formation rates are far lower, and many communities have seen a net decline in the number of locally owned businesses. This geographic stratification has become a political and economic fault line, one that policymakers are only beginning to address.

[IMAGE: Heatmap of U.S. county-level business formation rates, highlighting the concentration in coastal metros.]

Where Innovation Thrives: Sectors Defying the Trend (AI, Biotech, Clean Energy)

If the aggregate data on dynamism are gloomy, the innovation headlines are dazzling. Three sectors in particular capture the paradox: artificial intelligence, biotechnology, and clean energy.

Artificial Intelligence. Venture capital inflows into AI have exploded, reaching over $50 billion annually by 2024. Generative AI, large language models, and autonomous systems are spawning a new generation of startups. Yet the landscape is dominated by a handful of incumbents—Google, Microsoft, Meta, Amazon, and a few others—that control the underlying compute infrastructure, foundational models, and massive datasets. The result is a high-innovation, low-survival environment. Hundreds of AI startups are launched each year, but most either fail or are acquired quickly. The ones that succeed often become “born scale” companies, needing enormous capital and user bases from day one—a barrier that favors big players.

Biotechnology. The pace of novel drug approvals has remained strong, with the FDA greenlighting an average of 50 new molecular entities per year over the past decade, a 60 percent increase from the 2000s. CRISPR-based therapies have moved from lab to clinic, and mRNA platform technology opens the door to a new class of vaccines and treatments. But biotech is capital-intensive and characterized by long development cycles. Many breakthrough discoveries originate in university labs or small startups, only to be acquired by large pharmaceutical firms before reaching the market. The startup-to-scale pipeline is narrow, and the exit path is typically M&A rather than independent growth. This pattern means that innovation output remains high while the number of independent new firms that survive to maturity remains low.

Clean Energy. The Inflation Reduction Act (IRA) has triggered a surge in clean-tech investment, with over $300 billion in announced projects since 2022. Solar installations are at record levels, battery storage capacity is doubling annually, and electric vehicle sales are accelerating. Yet again, the value capture is concentrated. Large incumbents—energy utilities, automotive manufacturers, and established solar developers—are the primary beneficiaries. Smaller startups face formidable scaling challenges: they need to navigate permitting, grid interconnection, supply chain logistics, and capital markets that often favor established players. A handful of high-profile clean-tech startups have gone public, but many more have been acquired or folded.

Common Pattern. Across all three sectors, innovation is increasingly “deep-tech” and capital-intensive, favoring incumbents with deep pockets and long time horizons. The democratization of entrepreneurship that characterized the software boom of the 1990s and 2000s—where two people with laptops could launch a global business—is giving way to an era where breakthroughs demand specialized lab equipment, massive compute clusters, or regulatory expertise. This structural shift explains why aggregate dynamism has declined even as innovation appears to accelerate.

[IMAGE: Side-by-side comparison of startup survival curves for AI vs. traditional software (e.g., 1985 vs. 2023).]

The Double-Edged Sword of Market Concentration

Market concentration is often cast as a villain in the dynamism narrative. High concentration correlates with higher prices, lower wages, and reduced competition. But the relationship with innovation is more nuanced. In some industries, concentration suppresses innovation by allowing dominant firms to coast on their market power. In others—particularly those requiring heavy upfront R&D—concentration may actually foster innovation by enabling firms to capture the full returns from their investments.

Consider the case of semiconductor manufacturing. Fewer than five firms now control the most advanced fabrication capacity. The cost of building a leading-edge fab has surpassed $20 billion. This is not a market where a garage startup can compete. Yet the concentrated structure has not prevented a relentless pace of innovation: Moore’s Law continues, albeit at a slowing clip, and companies like TSMC and Intel are racing to push the boundaries of chip design. The question is whether this concentrated structure is stable and whether it denies opportunities to new entrants.

The answer likely varies by sector. In digital platforms, network effects create natural monopolies that can be innovation-hostile once dominance is established. In biopharma, the concentration of R&D capacity among a few large firms coexists with a vibrant ecosystem of academic labs and venture-funded startups that supply the pipeline. The policy challenge is not to indiscriminately break up large firms, but to identify where concentration has become a bottleneck and where it is a necessary condition for capital-intensive R&D.

[IMAGE: Diagram showing the innovation cycle: startups → acquisitions/partnerships → incumbents scaling, with policy intervention points highlighted.]

Policy as a Rebalancing Tool

If the paradox of declining dynamism and surging innovation is driven by structural rather than cyclical forces, then policy must respond accordingly. The traditional toolkit—tax incentives, patent reform, antitrust enforcement—needs updating.

Antitrust and competition policy. The Biden administration’s aggressive antitrust stance, led by the FTC and DOJ, has targeted large tech firms with lawsuits challenging their market power. The goal is to lower entry barriers. But the evidence that current antitrust enforcement has meaningfully increased dynamism is mixed. A smarter approach may focus on ensuring interoperability, data portability, and platform openness rather than breaking up companies. For example, requiring large platforms to allow third-party access to data could reduce the advantages of incumbency while preserving scale economies.

R&D and innovation funding. The CHIPS and Science Act, passed in 2022, directs billions into semiconductor manufacturing and R&D. The IRA does the same for clean energy. These are down payments on a strategy that recognizes the capital-intensive nature of modern innovation. But they must be complemented by policies that help startups cross the “valley of death” from lab to market. Programs like ARPA-E have shown that well-designed government funding can de-risk early-stage technologies and attract private capital.

Regional rebalancing. The geographic concentration of innovation is a major drag on national dynamism. Policies that expand access to venture capital, technology transfer, and workforce training in underserved regions could help. Place-based initiatives, such as the Economic Development Administration’s “Tech Hubs” program, aim to create innovation clusters outside the coasts. Early evidence is promising, but scaled implementation remains a challenge.

Immigration and talent. The U.S. innovation system is fueled by foreign-born talent. Approximately half of all billion-dollar startups have immigrant founders. Yet the immigration system remains restrictive, particularly for highly skilled workers. Expanding H-1B caps, creating a startup visa, and streamlining green cards for STEM graduates would directly boost dynamism.

[IMAGE: Map of proposed U.S. Regional Technology Hubs under the CHIPS Act, with existing tech hubs in coastal metros.]

Global Implications: Supply Chains, Competitiveness, and the Next Decade

The paradox is not only a U.S. story. Similar patterns—declining startup rates alongside accelerating deep-tech innovation—are visible in Europe, China, and other advanced economies. But the implications are most consequential for American competitiveness.

If the U.S. can sustain its innovation output while addressing the decline in dynamism, it will maintain its technological edge. If, however, the concentration of innovation among a handful of firms and regions leads to political backlash, regulatory overreach, or a failure to broaden the innovation pipeline, the long-term consequences could be severe. The supply chain vulnerabilities exposed by the COVID-19 pandemic and the semiconductor shortage have already prompted a national conversation about re-shoring critical technologies. Yet the deep-tech startups that will produce the next generation of semiconductors, batteries, and biomanufacturing systems need support to scale domestically.

For the global economy, the U.S. experience offers a cautionary tale: the metrics that traditionally measure economic health may be missing the real action. Policymakers around the world should not assume that falling startup rates signal inevitable decline. Instead, they should invest in understanding the new structural realities—and craft policies that encourage innovation without sacrificing the competitive churn that has long been the engine of prosperity.

[IMAGE: Global comparison chart: startup density vs. R&D output (patents per capita) for U.S., EU, China, and Japan.]

Conclusion: The Path Forward

The Great Rebalancing describes a transition, not a crisis. The U.S. economy is moving from a model of broad-based, low-barrier entrepreneurship to one of concentrated, capital-intensive, deep-tech innovation. This shift brings risks and opportunities. The risks include rising inequality, regional divergence, and the potential for entrenched incumbents to slow innovation once they no longer fear disruption. The opportunities include the possibility of solving major challenges—climate change, disease, and economic insecurity—through the intensified application of science and technology.

Navigating this paradox requires humility. The policy tools that worked in the 20th century may not fit the 21st. Antitrust enforcement that treats all concentration as harmful may miss the mark. Conversely, laissez-faire policies that let winners take all may ossify the economy. The path forward lies in a nuanced strategy: foster competition where feasible, invest in public R&D where markets underinvest, expand the geographic and demographic base of innovation, and ensure that the benefits of breakthrough technologies are broadly shared.

The next decade will test whether the United States can reconcile its two faces—the one showing a slow decline in dynamism, the other a surge in innovation. The answer will depend less on any single policy and more on whether the country can adapt its institutions to the structural realities of a transformed innovation landscape.