Why Europe produces world-class technology but too few technology empires — and why the usual Silicon Valley explanation gets the problem wrong.

Walk through Veldhoven and you encounter one of the strangest facts about the modern digital economy.

Nearly every advanced AI accelerator, smartphone processor and high-end server chip ultimately depends on machines built by a Dutch company. ASML's extreme-ultraviolet lithography systems sit close to the physical limit of what industrial engineering can accomplish. In 2025 ASML generated €32.7 billion in revenue and spent €4.7 billion on R&D. Its EUV technology remains critical to manufacturing leading-edge semiconductors.

Travel north and Spotify has more than 750 million monthly active users. Go south-east to Germany and SAP remains the world's leading enterprise-application software vendor by its own reporting and major market surveys. Europe is not exactly a continent where everyone spends the afternoon discussing medieval masonry while somebody in California invents computers.

Yet open the list of companies controlling the general-purpose infrastructure of the digital economy and the picture changes abruptly.

Operating systems: Microsoft, Apple, Google.

Cloud: Amazon, Microsoft, Google.

Search: Google.

Social platforms: Meta, ByteDance if we extend the comparison to China.

AI foundation platforms: OpenAI, Google, Anthropic, Meta, Microsoft-backed ecosystems.

Semiconductor design platforms and accelerators: NVIDIA, AMD, Qualcomm and increasingly specialised American firms.

The striking European absence is real.

But describing it as an innovation gap is too crude.

The more useful diagnosis is a compounding gap.

Europe is surprisingly good at creating knowledge, engineering difficult technology, forming startups and producing products people want. It is substantially worse at turning those achievements into corporations that accumulate capital, distribution, customers, acquisitions, infrastructure and political-economic gravity for twenty or thirty years.

That distinction matters enormously.

Because if the problem is that Europeans lack imagination, the prescription is to somehow manufacture more entrepreneurs.

If the problem is excessive taxation, the prescription is lower taxes.

If the problem is regulation, the prescription is deregulation.

And if the problem is that Europeans simply lack the magnificent hormonal profile of the Californian founder, presumably Brussels needs more hoodies, protein powder and podcasts.

The evidence points somewhere much less theatrical.

Europe does not primarily have an invention problem. It has a system for repeatedly interrupting the process between invention and durable market power.

That is both worse and better news.

Worse, because there is no single lever.

Better, because institutions can be changed.

First, let's kill the lazy version of the argument

The claim that Europe simply "doesn't innovate" does not survive contact with the data.

The 2025 WIPO Global Innovation Index ranked Switzerland first in the world, Sweden second and the United States third. The United Kingdom was sixth, Finland seventh, the Netherlands eighth and Denmark ninth. Germany ranked eleventh and France thirteenth. Europe supplied 15 of the world's top 25 innovation economies and six of the top ten.

Even inside the EU, which excludes innovation heavyweights Switzerland and Britain, the picture is hardly a technological wasteland. WIPO identifies Helsinki and Eindhoven among the world's most innovation-intensive clusters, followed by Stockholm, Copenhagen, Munich and Gothenburg. What differentiates American clusters most strongly is not the absence of European science but the much greater intensity of venture capital around American clusters.

The European Commission's own 2025 examination is similarly awkward for the "Europeans don't start companies" narrative. It estimated more than 35,000 early-stage startups in Europe and a technology workforce of around 3.5 million, while noting roughly comparable recent workforce growth to the United States. European technology firms had attracted about $426 billion in venture capital since 2015.

These numbers do not prove Europe has no innovation problem whatsoever. America remains exceptionally strong. WIPO finds that the United States leads globally in corporate R&D, unicorn valuation, software expenditure and intangible-asset intensity. American clusters combine research, venture financing and commercialization better than most European counterparts.

But they do destroy a simpler hypothesis:

Europe's shortage of giant technology companies cannot be explained merely by a shortage of inventions, scientists, engineers or startups.
Something happens later.

That "later" is where the interesting economics begins.

Mario Draghi framed Europe's predicament with one brutal statistic: when his competitiveness report was presented in 2024, no EU company worth more than €100 billion had been created from scratch during the previous fifty years, while all six American companies then worth more than €1 trillion had been founded during that period. He also noted that close to 30% of European-founded unicorns between 2008 and 2021 moved their headquarters abroad.

The exact €100 billion formulation has edge cases — corporate histories are messy creatures, and market capitalisation is hardly a divinely ordained measurement of human progress. But the underlying pattern is difficult to dispute. Only a handful of the world's dominant new digital corporations are headquartered in the EU.

What happened to European innovation instead?

Sometimes it became a successful European firm.

Sometimes it became an extremely successful European firm. ASML is a market leader of extraordinary strategic importance. SAP is a global enterprise-software champion. Spotify created a global consumer platform at genuine scale. Europe also retains formidable positions in telecom equipment, industrial automation, payments, aerospace, semiconductor equipment and scientific instrumentation.

But another recurring path is equally important:

European invention → European startup → successful product → foreign growth capital → foreign acquisition or foreign listing → incorporation into a larger non-European platform.

Skype illustrates the pattern almost too neatly. The European communications pioneer had 170 million connected users when Microsoft agreed to buy it for $8.5 billion in 2011. Microsoft later acquired Swedish game developer Mojang. PayPal bought Sweden's iZettle for about $2.2 billion after iZettle had been preparing for a Stockholm listing.

None of these transactions was necessarily bad for the founders.

This point is essential.

Selling a company for billions is not entrepreneurial failure. It would be slightly deranged to tell someone who has created several thousand jobs and then received a $2 billion acquisition offer that they have let down Charlemagne.

But what is rational for a founder can still produce an unfortunate macroeconomic pattern.

When a company is absorbed into Microsoft, Google or PayPal, the acquiring company receives more than some source code.

It acquires:

  • engineers;
  • intellectual property;
  • customers;
  • product knowledge;
  • management talent;
  • datasets;
  • distribution;
  • and sometimes the embryonic platform that might otherwise have become a competitor.

That makes the next acquisition easier.

And the next one.

This is what compounding means.

This is true — but only after we state it correctly.

Europe does not lack money.

It has enormous savings.

What Europe lacks is a financial system that efficiently turns enough of those savings into long-duration, high-risk equity financing for rapidly growing companies.

The distinction is almost comic.

Europeans save money, European institutions aggregate it, and a considerable amount of it eventually helps finance American technology companies.

ECB research published in 2026 estimated that roughly half of euro-area household equity exposure was to securities issued outside the EU. American equities alone represented about 34%, roughly the same as households' exposure to domestic equities. The ECB also found that euro-area holdings of US equities had quadrupled over the preceding decade.

So Europe has capital.

It merely possesses a financial plumbing system with a peculiar tendency to deliver some of that capital to Palo Alto.

The scale-up numbers are even more revealing. EIB research finds that by around their tenth year, EU scale-ups have raised approximately 50% less capital than comparable San Francisco firms. More than four out of five EU scale-up financing deals involve a foreign lead or sole investor, compared with about 14% in San Francisco. Annual US venture investment is six to eight times European levels.

This is not primarily a seed-funding problem.

Europe has become reasonably capable of financing birth.

It has trouble financing adolescence.

And technology companies are unusually hungry adolescents.

A software platform may require years of losses while it builds network effects. A semiconductor company may need billions before production. Foundation AI adds compute bills that resemble national infrastructure projects performed by people wearing trainers.

A company that can raise €20 million but cannot efficiently raise €500 million does not possess 4% of the financing system it needs. At some point it changes ecosystem.

It goes where the capital is.

This explanation is popular because it explains almost everything without requiring anyone to examine a securities-settlement system.

It is also suspiciously circular.

Why do Americans invest more aggressively in startups?

Because Americans are more risk-taking.

How do we know they're more risk-taking?

Because they invest more aggressively in startups.

A small snake has now swallowed its own cap table.

Institutions matter.

ECB work on venture-capital investors points to the historical role of US pension reforms in allowing greater allocations to risk assets. European venture financing, by contrast, remains unusually dependent on government-backed institutions. The ECB describes the European Investment Fund and public entities as central nodes in the European VC ecosystem, whereas American institutional markets provide a broader private capital base.

The scale of Europe's public role is revealing. In 2025 the European Investment Fund accounted for almost a quarter of venture capital raised by European funds, while the broader EIB Group supplied around 30% of EU venture debt.

That support is valuable.

But it also tells us something.

Europe's problem is not that its governments interfere with an otherwise perfect venture market. Quite often government institutions are keeping the venture market functioning.

The missing piece is a sufficiently deep private institutional layer behind them.

This theory has an empirical nuisance named Sweden.

Sweden ranks second in WIPO's global innovation index. Finland is seventh. Denmark ninth. The Netherlands eighth. These are not exactly jurisdictions famous for replacing public services with laminated copies of Atlas Shrugged.

Sweden produced Spotify, Klarna, Skype, Mojang and a continuing succession of technology companies. Estonia, with its European social model intact, developed one of the world's densest startup cultures relative to population.

Taxes matter at the margin. Employee stock-option treatment matters considerably. So do immigration rules and the administrative portability of benefits. The Commission's own scale-up analysis identifies stock-option taxation, social-security mobility and talent rules as specific European problems.

But those are mechanisms, not morality plays.

A Swedish engineer does not become genetically cautious at the Øresund Bridge.

Change the incentives and legal mechanics and behaviour changes.

The more interesting European weakness is therefore not "high taxes" in the abstract. It is that twenty-seven national systems often tax stock options differently, administer insolvency differently, structure corporations differently and move employee benefits differently.

The friction arrives one paper cut at a time.

Eventually the company notices it is bleeding.

This is the most fashionable explanation and perhaps the least historically serious.

GDPR came into effect in 2018.

The AI Act arrived far later.

Europe's shortage of Microsoft-, Apple- and Google-scale digital platforms was already obvious long before either existed.

A theory cannot convincingly explain an outcome that predates its alleged cause by two decades. That is basic temporal hygiene.

Regulation can absolutely damage innovation. Badly designed rules can impose high fixed costs on small companies, discourage experimentation, increase uncertainty and favour incumbents capable of employing battalions of compliance specialists.

But "regulation" is too imprecise a variable.

The central European problem is often fragmentation of regulation, not merely its severity.

A startup expanding from California to New York remains within one federal corporate and capital-market architecture.

A startup moving from France to Germany, Spain, Italy and Poland encounters a single market in theory but continuing differences in taxation, insolvency, labour administration, corporate law, securities practice and sector regulation.

The Commission's own 2025 scale-up analysis lists regulatory uncertainty and fragmentation, limited cross-border expansion and inconsistent national frameworks among the major obstacles. The ECB reaches the same conclusion about capital markets: Europe still does not possess a genuinely unified financial market.

This gives us a more useful distinction:

Strong common rules can sometimes increase scale. Twenty-seven slightly different sets of rules almost always decrease it.
There is nothing inherently startup-friendly about regulatory ambiguity.

A single European privacy rule may be easier for a company than twenty-seven privacy regimes.

A single European corporation would be easier than twenty-seven subtly different incorporation systems.

The problem is not simply how much Europe regulates.

It is whether Europe regulates as Europe.

One can believe this only by overlooking a remarkable amount of American history.

ARPANET came from ARPA.

TCP/IP emerged from DARPA-sponsored work involving Robert Kahn and Stanford's Vint Cerf.

DARPA helped miniaturise GPS receivers.

Federal support contributed to semiconductor-development programmes including SEMATECH.

The American defence-research system supported technologies that later became central to civilian computing, networking, navigation and semiconductor industries.

Even Douglas Engelbart's famous demonstration of hypertext, windows and the computer mouse grew from work backed by ARPA and the US Air Force.

Silicon Valley is not proof that governments should stay away from technology.

It may be one of history's better demonstrations of what happens when public research, mission-driven procurement, universities, private risk capital and large markets reinforce one another.

That last phrase matters.

The US state did not merely issue research grants and then wish everybody the very best.

Defence and space agencies became demanding early customers.

Procurement generated markets.

Technical programmes created networks of researchers, suppliers and firms.

The private sector then performed the extraordinarily important work of commercialization, productization and scale.

This is not central planning.

Nor is it laissez-faire.

It is an ecosystem.

Europe has often been rather good at funding the first half and strangely hesitant about the second.

European public procurement represents roughly 15% of GDP, making government one of the largest potential technology buyers on Earth. Yet the Commission acknowledges that innovation procurement remains underused compared with other economies. A 2025 European expert report found that public procurement frequently produces pilots without subsequent large-scale adoption; major obstacles included administrative risk aversion, structural exclusion of startups and weak institutional mechanisms for turning experiments into purchases.

This deserves more attention than another conference about entrepreneurial culture.

Imagine two companies.

Both have developed a novel cybersecurity platform.

The American company wins a significant federal agency contract. That reference customer validates the product, funds deployment, attracts engineers and tells private investors that somebody with a very serious threat model considers the technology real.

The European competitor wins a €400,000 research pilot with three municipalities, writes a 76-page deliverable, attends a closing workshop and is congratulated for European collaboration.

Two years later, everyone is confused about the valuation gap.

Europe spends enormous amounts of public money.

The weakness is often conversion of public expenditure into scalable demand.

The Commission now explicitly wants greater startup access to procurement and a stronger innovation-procurement regime. That diagnosis is right.

Europe doesn't merely need more venture capitalists.

It needs more venture customers.

The EU population exceeds that of the United States.

On a PowerPoint slide, this gives European startups an enormous home market.

On an invoice, things become less philosophical.

The Commission's 2025-26 agenda continues to describe the single market as incomplete, particularly in services, finance, energy and telecommunications. The EU is now pursuing a single-market roadmap, deeper capital integration and a "28th regime" intended to let innovative firms operate under a common European company-law framework.

This gets much closer to the central problem.

Digital companies exhibit strong increasing returns.

Once software has been developed, serving customer number ten million is often vastly cheaper than acquiring the first thousand. Platforms benefit from network effects. Large customer bases generate data. Data improves products. Revenue finances infrastructure. Infrastructure reduces unit cost. Lower unit cost permits expansion.

Scale does not merely make a technology company bigger.

Scale can make it better at becoming bigger.

The United States gives companies a large relatively integrated market in which to start that process.

Europe nominally does too.

But every national friction weakens the exponent.

That difference compounds for twenty years.

Then everyone looks at Apple's market capitalisation and decides Americans must have superior motivational books.

Founders and venture investors eventually need liquidity.

This is another place where American scale compounds itself.

The European Commission reports stock-market capitalisation at roughly 73% of EU GDP, compared with around 270% in the United States. European investment funds are much smaller, and market infrastructure remains exceptionally fragmented.

ECB analysis counted hundreds of European trading venues and dozens of national clearing and settlement institutions, while US infrastructure is dramatically more consolidated. US large-cap and mid-cap shares consequently enjoy greater trading liquidity, improving their attractiveness to both issuers and investors.

This creates a feedback loop:

deeper public market → better exit → larger VC funds → larger growth rounds → larger companies → more attractive public market.

Europe runs the same loop in reverse.

Smaller growth funds make American capital more attractive.

Foreign capital makes a US expansion more attractive.

A US listing becomes logical.

A US acquisition becomes attractive.

Europe loses another potential anchor company.

Then pension and investment funds complain that there are too few large European technology companies to invest in.

So they buy NVIDIA.

The system is almost elegant.

Suppose Europe and America produce identical breakthroughs tomorrow.

The subsequent paths are not symmetrical.

An American startup can potentially receive:

  • cloud credits and infrastructure from hyperscalers;
  • capital from exceptionally deep venture markets;
  • engineers who previously scaled global platforms;
  • distribution through operating systems and app ecosystems;
  • acquisitions by companies with hundreds of billions in cash flow;
  • and an eventual listing into the world's deepest technology equity market.

A successful ecosystem therefore creates the inputs required to create its successor.

Former employees become founders.

Early employees become angels.

Executives become board members.

Large companies become customers.

Customers become distribution channels.

Acquirers recycle founder wealth into new companies.

This is the famous startup "mafia" phenomenon, but the underlying mechanism is broader than social networks.

It is institutional memory plus capital.

Europe has increasingly developed these flywheels in Stockholm, London, Paris, Berlin, Amsterdam and Tallinn. But they remain smaller and more geographically fragmented.

This is one reason losing a potential market leader matters beyond national prestige.

You don't merely lose one headquarters.

You may lose twenty years of second-order companies.

Another popular chart shows American corporations spending far more on technology R&D.

That is true.

WIPO estimates US companies accounted for approximately 47% of global corporate R&D expenditure in its latest dataset, compared with about 22% for Europe. Software and ICT services are also unusually R&D-intensive sectors.

It is tempting to conclude:

America spends more on innovation, therefore it generates more technology companies.

Causality also runs the other way.

America possesses Microsoft, Alphabet, Amazon, Meta, Apple and NVIDIA.

These are gigantic corporations with gigantic R&D budgets.

Europe's corporate base remains disproportionately concentrated in automobiles, pharmaceuticals, industrial engineering and other mature industries.

So Europe's lower aggregate corporate digital R&D is partly the consequence of failing to build huge digital firms in previous cycles.

Tomorrow's R&D budget is produced by yesterday's market leader.

Again, compounding.

ASML is sometimes treated as the exception that proves Europe's decline.

I think it proves something more interesting.

Europe can build a nearly irreplaceable technology leader when five conditions coincide:

  1. extraordinarily difficult science;
  2. long investment horizons;
  3. dense supplier relationships;
  4. sophisticated industrial customers;
  5. institutional patience.

ASML's EUV programme took roughly a quarter-century of sustained engineering. Its 2025 work included a demonstration of a 1,000-watt EUV source, itself the result of accumulated optical, plasma, laser and manufacturing expertise.

This is not startup mythology.

There was no weekend MVP for extreme ultraviolet lithography.

There was physics.

And thirty years.

Europe is unusually strong in precisely these areas: industrial systems, advanced manufacturing, energy technology, scientific instrumentation, aerospace, pharmaceuticals and complex business processes.

The mistake would be concluding that Europe should surrender consumer software, cloud and AI platforms and become merely an exceptionally sophisticated supplier to everybody else's digital empire.

That is the comfortable version of decline.

ASML should instead reveal where Europe's next opportunity lies:

combine Europe's physical and institutional assets with software and AI before someone else builds the platform layer above them.

The generative-AI transition creates the first genuine architectural reset in computing since mobile and cloud.

And resets matter because incumbency briefly becomes less decisive.

In 2026 Europe has significant AI ingredients:

  • exceptional universities and research groups;
  • industrial customers with highly valuable proprietary data;
  • companies such as Mistral;
  • semiconductor equipment leadership through ASML;
  • considerable high-performance computing infrastructure;
  • and nineteen EuroHPC AI Factories, supplemented by thirteen associated antennas, intended to give startups and SMEs access to compute and expertise.

The EU is also moving toward AI Gigafactories, with €20 billion intended to support large-scale facilities and a broader InvestAI ambition to mobilise €200 billion.

Useful.

But compute infrastructure alone will not create European Googles.

A supercomputer is not a market.

The central 2026 question is whether Europe repeats its historical sequence:

fund research → produce excellent technology → establish promising startup → celebrate funding round → sell or relocate during scale-up.

Or whether it finally carries companies through the entire chain.

AI actually makes the difference more urgent.

Code generation is reducing software-development costs.

Automatic translation is reducing language barriers.

Models can accelerate internationalisation, customer support and localization.

Some traditional disadvantages of a multilingual continent are therefore declining.

But other constraints become more important:

  • compute;
  • power;
  • capital;
  • data rights;
  • distribution;
  • procurement;
  • and access to customers.

The next European policy failure would be building seventeen AI programmes without fixing the market into which their products emerge.

The outcome is not predetermined.

Think of three scenarios.

Scenario 1 — The laboratory continent

Europe remains extraordinarily innovative.

European universities publish world-class research.

European engineers develop clean-energy systems, robotics, quantum technologies, advanced materials and AI applications.

European governments subsidise pilots.

Then the most commercially promising companies raise Series C and D rounds from US investors, establish larger US operations and eventually list in New York or sell to American incumbents.

Europe becomes something like a global R&D department.

This world is not poor.

It might even be pleasant.

But it is strategically dependent.

The continent creates much of the value and captures a smaller fraction of the compounding return.

Scenario 2 — Fortress Europe

Europe correctly identifies dependency but chooses the wrong remedy.

Foreign technology is restricted.

National champions receive protection.

Procurement becomes protectionist rather than demanding.

Competition declines.

Local firms learn to satisfy government programmes rather than customers.

Europe gets sovereign products nobody voluntarily buys outside Europe.

This is also failure.

A protected market leader is not necessarily a market leader.

It may simply be a bureaucracy with an API.

Scenario 3 — The compounding continent

Europe keeps its openness but changes the mechanics of scale.

A company incorporates once under an EU-wide regime.

Employee equity travels across borders without absurd tax surprises.

One financing round can efficiently involve investors across the Union.

Pension and insurance capital allocates a modest but meaningful share to European growth equity.

European public authorities become serious first customers for European technology while maintaining competitive procurement.

A successful company can raise €500 million in Europe.

Then €2 billion.

Then list into a liquid European capital market without accepting a structural valuation penalty.

Acquisitions still happen — they should — but selling to America becomes an option rather than the default endgame.

Successful founders reinvest.

Managers circulate.

New firms emerge.

Then the exponent changes.

This third scenario does not require Europeans to become Americans.

That may be its most important property.

The prescription follows naturally from the diagnosis.

1. Finish the single market for companies

The proposed EU Inc./28th regime is potentially more important than another subsidy programme. In March 2026 the Commission advanced the initiative as part of its Startup and Scaleup Strategy, aiming at common company-law rules for innovative firms.

Push this much further.

A genuine European startup should be able to operate across the Union with one corporate identity, portable employee equity, predictable insolvency and substantially harmonised administrative requirements.

A market of 450 million people must behave like one before we congratulate ourselves on having one.

2. Turn the Savings and Investments Union into actual equity

Europe does not need to persuade every household to invest its electricity bill in biotech startups.

It needs institutional plumbing.

Pension funds, insurers, endowments and investment funds need efficient routes into diversified European growth equity.

The ECB explicitly identifies deeper private markets and greater institutional involvement as necessary for financing innovative companies.

The objective is not nationalism in portfolios.

It is correcting a system in which European savings can finance American innovation more efficiently than European innovation.

3. Build a continental growth-capital layer

The newly established €5 billion Scaleup Europe Fund is a sensible intervention, and as of August 2026 the Commission had completed the legal steps required for it to begin investing.

But €5 billion is scaffolding.

It cannot substitute for a functioning private market measured in hundreds of billions.

Use public capital to create funds, expertise and track records that eventually crowd in pension, insurance and sovereign capital.

The end goal should be that the public fund becomes less necessary because it succeeded.

4. Turn procurement into a technology policy

European governments should behave more like intelligent lead customers.

Not "buy European regardless of quality."

That breeds expensive mediocrity.

Instead:

identify difficult problems, fund competing solutions, purchase successful ones at meaningful scale.

That is far closer to the productive American DARPA tradition than either grant-making or protectionism. Europe already recognises innovation procurement as underused and is developing reforms.

A €10 million commercial contract is often more transformative to a startup than a €10 million innovation grant.

One produces paperwork.

The other produces a customer.

5. Stop measuring success at the funding round

European technology policy still celebrates startup formation and venture rounds disproportionately.

Those are inputs.

Measure instead:

  • European firms exceeding €10 billion revenue;
  • firms retaining European headquarters after ten years;
  • €500 million+ growth rounds financed primarily in Europe;
  • IPOs remaining on European markets;
  • global market share;
  • cross-border European revenue;
  • public procurement awarded to younger innovative firms;
  • and the number of second-generation startups created by employees of earlier European scale-ups.

That tells us whether an ecosystem is compounding.

6. Make competition policy create openings for entrants

European regulation is potentially an advantage when it increases contestability.

Interoperability, data portability and restrictions on incumbent self-preferencing can reduce the distribution advantage of existing platforms.

But regulation should be evaluated like software.

Does it accomplish the desired result?

What does it cost a five-person company?

Can compliance be automated?

Are twenty-seven regulators interpreting it consistently?

Delete rules that fail those tests.

Keep rules that make markets work.

This is neither deregulation nor regulatory maximalism.

It is engineering.

There is much Europe should learn from Silicon Valley.

Speed matters.

Equity matters.

Tolerance of failure matters.

Ambitious founders matter.

Deep technical networks matter.

But mythology confuses those emergent properties with their causes.

Silicon Valley is not successful because its residents discovered capitalism while Europeans were having lunch.

It is the result of an unusual combination:

federal research money + defence procurement + exceptional universities + immigration + one huge market + deep equity markets + stock-based compensation + repeated successful exits + technology incumbents + reinvested founder wealth.

Remove half of those and asking Europeans to "take more risks" becomes motivational theatre.

Europe should copy the mechanisms, not the costume.

No compulsory Patagonia vest is necessary.

So, does Europe have an innovation gap with America?

Yes.

But not quite the one usually described.

America is better at several forms of innovation, particularly digital commercialization, corporate R&D and venture-backed scaling. WIPO's data make that clear.

Europe's scientific and entrepreneurial foundations, however, are much stronger than its stock-market outcomes imply.

The larger weakness occurs between invention and institutional permanence.

It is a gap in:

  • market integration;
  • late-stage financing;
  • equity-market depth;
  • procurement;
  • exit options;
  • platform accumulation;
  • and reinvestment.

Call it the innovation-to-control gap.

Europe frequently creates technology.

America is much better at turning technology into institutions that control markets — and then using those institutions to finance the next technology.

That's why this argument matters beyond GDP.

Today's technology leaders determine tomorrow's infrastructure.

Cloud providers influence where data lives.

AI platforms influence which models companies use.

Operating systems define distribution.

Semiconductor platforms shape what can be computed.

Payment networks shape commerce.

The firms that compound eventually become part of the architecture of sovereignty itself.

The strangest part of the story is that many pieces of the solution already exist.

There is enough science.

There are enough engineers.

There are enough startups.

There is enough private wealth.

There is a huge market.

There are giant public procurement budgets.

There are world-leading industrial customers.

There are sophisticated states.

What is missing is the connective tissue.

A single market that behaves like one.

Capital that can remain with a company through scale.

Public institutions prepared to become customers rather than merely sponsors.

Liquid markets where successful companies can stay European without accepting a financial penalty.

And enough time for the best firms to become old.

That last point is underrated.

Microsoft was founded in 1975.

Apple in 1976.

Amazon in 1994.

Google in 1998.

NVIDIA in 1993.

Technology empires are not produced by a three-year innovation programme.

They accumulate.

Layer upon layer.

Customers upon customers.

Engineers teach engineers. Capital produces capital. Distribution produces distribution. The company gradually becomes a small universe with its own gravity.

Europe is very good at lighting stars.

Its problem is that too many leave the galaxy before they acquire enough mass.

The challenge for the next decade is therefore not to teach Europe how to innovate.

It already knows.

The challenge is to let European innovation compound.

And that is a considerably more hopeful problem to have.