Why Does the EU Have No Tech Industry?
A number of scholars and economists have debated this question for years, and offered very different proposals for why. Here are three leading theories and what they mean for EU tech and regulation.
It has been thirty years since the technology revolution began in Silicon Valley. And while California is still the epicenter of that boom, many other dominant online technology players have started (or moved to) different places around the United States in the intervening decades. In particular — the top 7 U.S. tech companies by market cap (MSFT, Alphabet, Amazon, Meta, Nvidia, Apple, Tesla) — are also in the top 10 tech companies in the world.
Those seven U.S. tech giants brought in $1,720 billion in revenue in 2023 . . . whereas Europe’s tech industry generated a mere $133 billion. To put that difference to scale, that is something the size of a basketball compared to the size of penny.
And in terms of market cap, Europe’s top seven tech companies have 20 times less value than those top tech companies.
What is remarkable about that difference is not how America achieved such dominance, but more interestingly, why — despite decades of effort — Europe’s tech industry is a mere shadow of the United States’.
Today’s newsletter summarizes three different theories of law, market, and social conditions that try to explain why Europe’s tech industry has failed to ever launch, and what that means for its future.
Anupam Chander: It’s the (lack of) law
One of the first papers I’m aware of with the necessary amount of perspective (written in 2013, leaving time for the phenomenon to reveal itself) is by Georgetown law professor Anupam Chander. In How Law Made Silicon Valley, Chander not only asks why not Europe (or Asia) but why specifically California’s Silicon Valley became a hot bed of innovation and entrepreneurialism. “[E]ducation and money coincide in other parts of the United States as well. Why did those parts no prosper in the manner of Silicon Valley?” he asks. “More fundamentally, why did the Internet not make geography irrelevant?”(641).
Chander argues that the primary reason for the creation of American dominance was “insulation brought by U.S. law reforms in the 1990s,” which essentially created a safe and relatively liability free place for the nascent tech industry “by both reducing the legal risks they faced and largely refraining from regulating the new risks they introduced” (645). It reminds me of the idea famously pointed out by Karl Polanyi that there is no such thing as a self-regulatory market: “regulation and markets, in effect, grew up together.”
Importantly, this is more than just a play on the typical tropes of the American versus European approach to regulation of markets (which says Europeans tend to regulate markets more, creating both practical hurdles to start-up innovation and less profit incentive to innovate). Instead, Chander is very specific: the U.S did more than just not impose new regulatory constraints on tech; it used law to make it easier for new technology companies to thrive. He gives three primary points of comparison between the U.S. and Europe, which I’ll quickly summarize:
(1) The U.S. created more favorable intermediary liability regimes
In the U.S. we can credit Congress’s passage of the §230 of the Communications Decency Act (CDA) which removed intermediary tort liability from websites in order to “simultaneously promote the speech potential of a largely self-regulated Internet, while fostering the rise of Internet enterprises”— as well as the Court’s broad reading of §230 to protect websites as both publishers and distributors (Zeran) and striking down the rest of the CDA, which would have created huge compliance obligations for companies (Reno v. ACLU).
In Europe, while the Electronic Commerce Directive offers some protection for liability for internet intermediaries “it stops far short of the near blanket exclusion” offered by §230. The result is the platforms face much more uncertainty (and cost) in knowing what they have to take down.
(2) The U.S. created more favorable copyright regimes
In the U.S., Chander credits the Digital Millennium Copyright Act’s Title II safe harbors for commercial enterprises; the ability for platforms to easily administer Copyright notice and takedown regimes; the Court’s expansive protection of internet services tools against copyright claims (Sony); and “the American concept of ‘fair use’” (664) with U.S. tech platforms early growth.
Unlike the U.S.. the EU’s E-Commerce Directive gives immunities from copyright liability, but does not give specifics as to what those might be. Chander argues the vagueness and lack of broad carve-outs of protection, creates uncertainty in the market and slows start-up investment.
(3) The U.S. basically has no privacy regimes
The U.S. has a very thin privacy regime. Common law privacy torts are incredibly narrow in scope and do little to address concerns around internet users’ protection of private information. Nascent tech companies thus have few obligations or potential risks in this regard.
In comparison, the EU was early to create privacy requirements for tech companies. In 1995, the Data Protection Directive was a large scale omnibus regulation that imposed obligations on all personal information. The 2002 E-Privacy Directive added even more privacy obligations to platforms. These not only increase the cost of compliance for platforms but the potential risks of operation, making starting a new tech company in a place like Europe much less attractive (682).
Anu Bradford: It’s everything else
It has been ten years since Chander spelled out the legal regimes that he argues helped to create America’s exceptionalism in tech innovation and industry, and in her latest article, Columbia law professor Anu Bradford pushes back on this narrative.
Or rather, she pushes back on what Chander’s arguments over the course of a decade have been often reduced to —the trope that I referred to before, that EU’s heavy-handed regulation has killed the viability of new markets and innovation and led to its anemic tech industry.
Bradford has written two books on comparative markets and technology (both of which are excellent). The first, The Brussels Effect, names the phenomenon of how the power and size of the European market, combined with the European Union’s ability to promulgate regulation has led to its outsized impact on global commerce and standards. In the U.S. we would call this the California Emissions Standards Effect — to describe how California’s large market size made it infeasible for car manufacturers to set different compliance regimes between states, resulting in U.S.-wide compliance with the California statute that heightened car emission requirements. There is also a normative argument coming out of The Brussels Effect, however: the idea that Europe’s role in raising global standards across multinational industries is a good thing not just for the EU but for the world.
I will leave discussion on that for another newsletter (I’m a fan of Bradford’s and her recent new book Digital Empires), but I offer it now in order to give perspective on Bradford’s priors and her rebuttal of Chander: she likes regulation and thinks that it is the way to both preserve Europe’s status as a super-power as well as to create better products and standards for consumers worldwide. Instead, in a new paper forthcoming in Northwestern University Law Review, she offers four other culprits to be blamed for the US-EU tech industry gap:
(1) Europe’s cultures and languages mean they can’t scale operations as easily as the U.S.
Though there is a European Union, there is much more variance in language, cultures and government regulations between member states which makes it harder to scale up business and thus for tech companies to get off the ground. As Bradford argues, “scaling is key to growth and competitiveness, yet such a growth strategy is harder to pursue when companies are operating across numerous national markets with different languages, cultures, and government regulations” (34).
It’s not just that there is regulation or that it’s strict. It’s that the regulation is hard to comply with — and that’s true with or without EU’s market wide laws. “The problem for tech companies is not often the regulatory stringency in Europe as much as it is regulatory complexity owing to the absence of common European rules. An alternative to a GDPR, AI Act, DMA, DSA, and other major European-level digital regulations is not a Europe without digital regulation; the alternative is a Europe with twenty-seven different digital regulations, adding to the complexity that is already hampering tech companies’ growth strategies in Europe. As a result, laws such as the GDPR are more likely to facilitate than undermine innovation, by mitigating uncertainty and complexity” (37).
(2) America has better capital markets to invest in innovation
Bradford argues that it’s not just that there are more financial intermediaries funding tech companies in the U.S. but it’s also the type of investors that provide the capital that VC firms deploy. Not only has the U.S. government played a more active role in funding U.S. tech innovation, but the presence of institutional investors such as universities and pension funds that invest in risky startups elevates opportunity for U.S. firms (39).
Addressing the obvious point that money doesn’t have borders when it comes to good opportunities for investment, Bradford acknowledges that often risky investment just tends to favor local companies because VC firms “feel more comfortable making risky bets on companies whose funders they know and whose business operations they can closely monitor after making the investment” (40). I’d call this the Silicon Valley VC Hallway Effect.
(3) Europe’s bankruptcy laws deter risk-taking
“Across several dimensions, US personal insolvency regimes are less punitive for the entrepreneur in case of failure, lowering barriers to entrepreneurship and risk taking” (42-43).
(4) Europe’s stringent immigration laws hurt their access to talent
Bradford posits that the innovation deficit in Europe is actually a migrant talent deficit — one that Europe could solve through a proactive migration policy. “In comparison, the U.S. technology sector relies heavily on its ability to attract immigrants. . . 55% of America’s billion-dollar companies have an immigrant founder. . . Overall, studies have documented that immigrants are more entrepreneurial than the general U.S. population” (46-47).
Plus, American companies pay better. “...foreign talent prefers the US because of the financial rewards available in the US for tech entrepreneurs and employees. A 2017 study by European VC firm Index Ventures found that more than 75% of EU countries’ stock-option rules lagged those in the US’ (51).
Why not both?
I came back to these papers and started thinking about this newsletter after reading a short essay published a few weeks ago by economist David Evans, which argues against many of Bradford’s and Chander’s claims, but not all of them. In reading the pieces — and after 9 months of sitting in Europe talking to tech workers and regulators here — I think there is more consensus between all these positions than disagreement.
Europe’s combination of cultural, language, and legal fragmentation absolutely do create an enormous hurdle for scale
In disagreeing with Bradford, language and culture fragmentation might explain some of Europe’s failure to innovate, Evans claims, but it can’t explain all of it. He gives the example of MalaGrab, a Malaysian leading company in deliveries, mobility, and payments that started in Malaysia in 2012, and now operates in Cambodia, Indonesia, Malaysia, Myanmar, Philippines, Singapore, Thailand and Vietnam. “These countries are at least as diverse and fragmented as the European ones.”
I think that’s probably right, but I also think it misses the second incredibly valuable point that Bradford raises alongside the culture and language issues: the difficulty in operating across Europe given regulatory fragmentation between countries while also complying with EU law. Add the language barriers to that legal compliance and this truly is a tremendous burden.
Germany’s administrative and regulatory regime alone is a notoriously dense and difficult to navigate, even for Germans: In February Germany missed the deadline to appoint a Digital Service Coordinator to comply with the DSA because of difficulties dividing responsibilities between its numerous regulatory bodies. How is a nascent tech business supposed to know how to start operating in Germany if they don’t know how European or German law might be applied to them?
More importantly why would a nascent tech business try to startup in that environment of uncertainty?
Law can explain a lot but not all of Europe’s failure to innovate
Building on that idea of uncertainty, I think this is the hidden importance of Chander’s three main arguments on copyright, intermediary liability, and privacy laws’ roles in building the U.S. tech industry. U.S. laws such that there were fewer of them and their construction by courts was explicit, broad, and relatively quickly created highly predictable legal regime in the U.S. tech industry. Predictability is good for markets and encourages investment.
Chander is also right in that the U.S. created favorable market conditions — many even call §230’s broad intermediary liability protection a type of subsidy to tech startups. But Bradford is also right that most of Europe’s big new omnibus regulation (GDPR, DSA, DMA) all came two decades into the EU failing to enter the tech markets, and so can’t have been the sole cause. Even Evans agrees on this point: “Greater regulation can explain why Europe has had a lower rate of formation of digital businesses, and, therefore, why relatively fewer home-grown businesses become digital leaders in Europe than the U.S. It is a less plausible excuse for Europe having created so few digital successes over three decades. Digital businesses, started elsewhere, come to Europe, often soon after they have started, and build up substantial European operations.”
Lower-risk American VC would invest long-distance in Europe if there were good ideas to invest in, but high-risk American VC still relies on tech hub networks that Europe doesn’t have and now might never have
Evans points out correctly that venture capital would come to the EU if the opportunities were there to invest in, and I agree with this point, but only for lower risk VC profiles. The Silicon Valley effect described by Chander has much more to do with what I called the Silicon Valley VC Hallway Effect — personal in person networking, “knowing a guy,” or just rubbing elbows with high net value individuals that are willing to invest a little or a lot in a person standing in front of them . . . and that comes from their community, university, or cultural world-view dramatically lowers the friction to investment. This combines three of Bradford’s ideas: the types of investors, the importance of in-person local networks, and also the talent pool that comes from American universities.
Europe should change its migration policies for lots of reasons, but it could also help tech
I strongly agree with Bradford’s prescription around innovation and tech, and that it was a contributing factor in lack of innovation in the tech space — but certainly not the only one. There are unwritten complicating factors here: migration and talent workforces are often generational endeavors — talented young people can come to the U.S. for a job and stay with other relatives who have travelled there already. American culture and educational system rewards ingenuity, entrepreneurialism, and diversity.
Migration is incredibly difficult and costly. Things have to either be really bad where you’re from, or really good where you’re going. Assuming we control for the former, the U.S. tech sector will always beat Europe on salaries, equity, or bonuses. If you’re going to move countries for a job, you might as well get paid. And if you’re going to move countries to start a company you might as well go to the country where the laws are most favorable to you and your start up.
I don’t know what it is, but I don’t think it’s bankruptcy law
In the end, I don’t know that it matters whether there is one thing (it’s never just one thing) that can explain Europe’s failure to innovate in tech these last thirty years, but I will say I don’t think it’s bankruptcy law. Though I appreciate Bradford’s point on this I do not think that most individuals are thinking about punitive insolvency laws before starting a business anymore than people think about a heart attack every time they have a Big Mac.
What I will say is that I agree with Chander on the unique role that early U.S. policy and common law played in fostering Silicon Valley, and with Bradford in the unique fragmented regulatory environment of the EU. I also think that she correctly points to differences between the U.S. and Europe in attracting, keeping, and educating talent is hand in glove with the failure to develop tech hubs and network effects that would increase high risk VC funding to Europe.
Evans essay ends with a call to action for Europe to learn from it’s past mistakes before “the next major technological disruption takes place, say 2040,” and Europe finds “itself in the same situation unless it takes actions that induce a high degree of entrepreneurship, investment, and risk-taking, which could lead to success.”
I don’t know that I put much hope in that particular market driven future for Europe. I find it much more likely that it will continue doing what Bradford thinks is the best bet for Europe’s future: driving regulatory action of global multinational corporations and creating new industries of compliance on that regulation. But that’s a newsletter for another day.
A serious problem in the analysis is to look at "Europe" and the "USA" only in terms of global behemoths. Scaling to the largest scale is easier done in the USA, because of the size of the domestic market, which in turn creates global visibility. However, Spotify, Booking, SAP, Just Eat, ASML, Ericsson, Nokia, Orange, Ubisoft, Telefonica are examples of European players with massive reach across the world. Amazon may be big in retail in the USA. In continental Europe it faces more difficulty, because of stronger local competitors, such as Coolblue in the Netherlands, Mediamarkt in Germany etc.
The capital market issue is a massive thing however. A tech startup in the Netherlands will have a really hard time getting tens of millions of funding. To have any chance in the USA that type of money is needed. Scaling to global level is then eeasier from the USA than from the Netherlands, Belgium or Germany.
Thanks for the careful read of my paper, Kate--and what a great post, canvassing and appraising Anu's and my approaches. I learned from your framing.