Adnkronos (Adnkronos) – Non è la fine della globalizzazione. È qualcosa di più profondo: una riscrittura della mappa delle dipendenze che per decenni abbiamo considerato soltanto come il prodotto efficiente della specializzazione internazionale. Un chip, una rete di pagamenti, un magnete al neodimio, un servizio cloud non valgono più soltanto per ciò che producono o per quanto costano, ma per il potere che attribuiscono a chi li controlla e per la vulnerabilità che creano in chi ne dipende.
È su questo confine, sempre meno netto, tra economia e geopolitica che lavora Matteo Maggiori, professore di Finanza alla Stanford Graduate School of Business, senior fellow allo Stanford Institute for Economic Policy Research e cofondatore del Global Capital Allocation Project.
La sua ricerca, nata dallo studio della finanza internazionale, dei flussi di capitale, delle valute di riserva e del ruolo del dollaro, negli ultimi anni si è spostata sempre più decisamente verso la geoeconomia: il modo in cui gli Stati trasformano relazioni economiche apparentemente ordinarie in strumenti di influenza e coercizione.
Con Christopher Clayton e Jesse Schreger, Maggiori ha costruito un’architettura teorica del fenomeno.
In The Great Game: A Model of Geoeconomic Competition, due egemoni continuano a commerciare tra loro mentre cercano contemporaneamente di impedire che il rivale diventi un’alternativa credibile per il resto del mondo: il profitto di una relazione economica viene così continuamente pesato contro il potere geopolitico che quella stessa relazione può consegnare all’avversario.
Maggiori è in questi giorni a Torino per il Festival Internazionale dell’Economia.
Al Collegio Carlo Alberto tiene la Pareto Lecture dedicata alla “Geoeconomics”: è da qui che parte anche la conversazione con l’Adnkronos.
There's constant talk of deglobalization, decoupling, and friend-shoring. Are we really witnessing the end of globalization?
No. It's a much more subtle change. If we look at international trade data, we don't simply see a world that's stopped trading: we're still at very high levels. What we see is rather a reassessment, a reconsideration of the nature of globalization and, above all, of economic dependencies.
For Europe, this is crucial. And the shock, in my opinion, didn't come only or even primarily from China: it came from the United States.
Let's take payment systems. They're a classic example of a system you want to be part of because everyone else is there, a bit like a social network. For retail purchases in Europe, we use Visa and Mastercard extensively. Internationally, we have Swift. A question that almost no one would have asked ten years ago is now natural: how dependent do we want to be on infrastructures that can be influenced by another country?
This is where initiatives like the digital euro and the attempt to further connect European wholesale payment systems with those of countries like Brazil and India come from. And the same reasoning applies to companies like ASML, which is located at a key hub in the global semiconductor supply chain.
For decades, we've explained the advantages of globalization through specialization: I do one thing, you do another, we both become very good at it, and then we trade. It's the classic economic argument. Now we need to add a question: what happens if one day the other party tells you they can't sell you that thing anymore, at any price?
Many productions can be recreated fairly quickly. Others cannot. You can't build an international payment system or advanced semiconductor manufacturing capacity overnight. This is where the trade-off between the advantage of specialization and the value of economic security emerges.
But will this really lead to the formation of two economic systems, one Western and one Chinese?
Here we must start with the importance of allies. The United States is a huge country and has very strong economic power, but in many areas, to exert real pressure on third countries, it needs Europe, the United Kingdom, Korea, and Australia.
Consider financial sanctions: without the major European and British financial centers, their effectiveness could be greatly reduced. Consider semiconductors: some essential technologies exist in Europe, while others exist in Korea. Exercising and maintaining geoeconomic power therefore requires a network of alliances.
Therefore, if America's goal is to compete with China, destroying the transatlantic alliance or introducing significant uncertainty into it is probably one of the least sensible economic policies that could be pursued.
At the same time, I don't see a simple disappearance of trade. The ratio of imports and exports to global GDP, after the significant increase in the decades before 2008, has essentially stabilized. And at this stage, the AI boom is actually fueling significant international trade, because its manufacturing and infrastructure are distributed across many countries.
So the structural change won't necessarily be less globalization. It will be a different kind of globalization: a rotation of activities, value chains, and dependencies."
The problem, however, is establishing what is truly strategic. Nowadays, any industry in difficulty tends to define itself as such.
And that's exactly the risk. Most industries aren't strategic. One of the reasons we're doing this research is precisely to try to discipline the concept. Otherwise, the mechanism becomes very simple: you're a European company that can't compete and you ask for protection by saying that if you close, the Chinese will come. But an economist is primarily interested in people, not in keeping any company afloat. If a closure causes a very rapid shock, concentrated on workers who have few alternatives, there are excellent reasons for support and retraining policies. It's different from saying that that specific company must survive.
For a sector to be truly strategic, I see at least three characteristics.
The first is a very tight control over the supply. If someone tells me, 'I won't sell it to you anymore,' I have to ask myself if there's another supplier. Controlling 95% of something is very different from controlling 50%.
The second is the technological difficulty of recreating the alternative. This is one of the reasons why ASML matters so much. Rebuilding certain capabilities requires years, technology, and expertise that can't be bought simply by putting money on the table.
The third is the sector's importance for the rest of the economy. Payment systems are strategic because blocking transactions has a huge cross-cutting effect. The same isn't true for the vast majority of goods we trade.
So we need to stop using 'strategic' as a binary category or a political label. It's a quantitative issue."
Let's take concrete examples: solar panels, wind turbines, batteries—in short, green energy. Should Europe accept buying much cheaper Chinese products or invest (and pay higher prices) in rebuilding its own industry?
"We first need to understand exactly what the dependency is. If China stopped selling us solar panels tomorrow, the ones we've already installed would continue to produce energy. It's different from oil: oil is consumed, and if they stop selling it to you, you only have available reserves.
This is important because the answer to dependency doesn't necessarily have to be: let's produce everything in-house. You can have strategic inventories. You can install greater capacity. You can diversify suppliers.
The transition to renewable energy, among other things, could also benefit Europe's economic independence. We don't have enough oil or natural gas, but we can produce wind and solar power. This is a good reason to build capacity.
But we must abandon the idea that Europe must be able to produce everything, or that the mere existence of a highly efficient foreign competitor automatically constitutes a geostrategic threat. This is said by someone who dedicates much of his research precisely to demonstrating the existence of certain geoeconomic dependencies. But there's a huge distance between this and arguing that every country should have its own steel mill.
So how should a European industrial policy work?
“The first principle should be: subsidize a sector, not a company.
If we decide that a certain technology is truly strategic, we can create incentives for European expertise to emerge in that sector. But then there must be intense competition among companies, and we must accept that the best one wins.
And this, politically, is difficult for Europe. Imagine we finance a strategic sector and in the end all the winning companies are German. How long will it take before we go back to saying: so we want an Italian one, a Spanish one, a French one?
At that point, we're back to the politics of national champions. And that policy didn't work.
In the United States, this problem is less acute. If the chip industry is financed and many factories end up in Texas, the political system can absorb it. In Europe, we need to be able to truly think about a European industrial policy, not twenty-seven national industrial policies coordinated through Brussels."
The issue with AI becomes even more delicate. Europe is talking about gigafactories, technological sovereignty, and its own stack. Is it realistic to chase the United States and China along the entire value chain?
“Here I would distinguish at least three things.
The first: we're dealing with a technology that's changing at an enormous rate. Every six months, we discover new uses and new possibilities. The idea that a public administration can decide from above which technology or which specific company will win seems rather crazy to me.
The government can create incentives, infrastructure, and fiscal and regulatory conditions that favor the sector. But innovation must occur through intense private competition.
The second distinction is between the training of frontier models and inference, that is, the use of models.
Training models at the frontier requires enormous investments and is, at least under current conditions, a very difficult objective for Europe. But much of AI's impact on productivity and the real economy, and a significant part of sovereignty issues, has to do with inference and therefore adoption. That costs much less.
Currently, there are also open-weight models, both American and Chinese, that can be transferred and used in Europe and remain relatively close to the border. We don't know if this will always be the case.
For this reason, a realistic European strategy could be to build a large inference infrastructure, ensure that data can remain in European data centers and, above all, encourage much more widespread adoption of AI.
It seems more sensible to me than deciding to create by decree a European company that can compete with the big American cloud providers and ending up, as De André said, with the State 'getting involved, getting indignant, and then throwing in the towel with great dignity.'"
However, the problem of computing capacity and data centers remains.
That's a real problem. But there's also a lot of relatively concrete economic policy to be done there.
We must make construction possible. The Letta and Draghi reports raised important questions about the integration of European markets. We cannot complain about lacking computing infrastructure while simultaneously making it nearly impossible to build a data center due to regulatory, energy, administrative, or labor issues.
Europe is big and has the capacity to do this.
And we must leave room for private innovation here too, because today we don't know what the optimal data center will be in two or three years. That risk must be assumed by the investor. The role of politicians is not to create conditions that prevent investment from occurring and then complain that the private sector isn't investing.
Let's talk about the Savings and Investment Union. Living in Stanford, just steps from the venture capital ecosystem, she identifies a much less talked about European obstacle: bankruptcy law.
This is a topic I care deeply about. The system here works by accepting that investing in a startup is extremely risky. Venture capitalists know that many companies will fail, and that only a few will pay off. When a project doesn't work, the company is quickly shut down, the remaining value is recovered, and the next investment is made.
It is a system that can also produce spectacular failures and fraud, but it also produces a great deal of innovation.
However, if in Italy a young person wanting to start a business has to put up their parents' house as collateral, the problem isn't simply that banks aren't willing to take the risk. If a company goes bankrupt and takes fifteen years to liquidate, destroying much of its value, it's reasonable for private capital to be much more prudent.
Therefore, it's not enough to say, "Let's unite the markets." Capital can take risks if there is also a system in which gains and losses can be realized quickly.
In Europe, we also come from two or three centuries of strong cultural and legal penalization of failure. In the American system, failure is part of trying: you try, it doesn't work, you move on.
I believe this is an issue Europe should seriously address. And if the conditions were right, it's not at all true that American venture capital wouldn't be interested in investing in Europe. There's no shortage of venture capital around the world. What's often lacking are the conditions under which it can be deployed.
And the risk of an American "switch-off"? Cloud, software, payment systems. Is this a realistic scenario, or would American multinationals have too much to lose?
The conflict between corporate interests and the state's geopolitical interests already exists. The most obvious example is Nvidia.
China is a huge market for Nvidia. When the Biden administration imposed restrictions on the export of advanced chips, Nvidia certainly wasn't happy. It's losing revenue, losing scale, and even running the risk of Beijing subsidizing a competitor that could one day become competitive.
The same goes for Asml: it's not as if the Dutch company had a spontaneous interest in not selling its machines in China.
So the conflict exists.
The Biden administration's approach was summarized by Jake Sullivan with the image of the small yard, high fence: a small garden surrounded by a very high barrier, meaning very few truly strategic technologies subjected to very strict controls.
In the end, jokingly, I'd say that instead of the small garden we put medium-high and somewhat fragile barriers around the entire Villa Borghese.
But above all, something has changed for the allies. The original idea was that Canada, the United Kingdom, or Europe shouldn't worry about these forms of American economic coercion. China, Russia, and Iran should.
Trump has profoundly changed that perception.
And paradoxically, I believe the shock has been greater for Europe and Canada than for China. In Beijing, no one is surprised by the existence of a geoeconomic competition with Washington. For Europeans, however, the uncertainty has affected their industries, tariffs, and even America's commitment to NATO.
Could it just be a parenthesis linked to an American administration?
“When you introduce uncertainty into commitments of this kind, the problem doesn't go away simply because the occupant of the White House changes.
Security policies have a twenty-year horizon. If you discover that a commitment you considered structural may be called into question, you must incorporate that risk into your future decisions.
From this point of view, I believe we will have to find a new balance and that simply returning to the previous one is not possible.
This doesn't mean the United States, Europe, and China will stop trading. Economic interdependence creates enormous value and therefore also imposes limits on coercion.
During the Cold War, trade between the United States and the Soviet Union was much more segmented. Today, telling China, "We won't sell you anything that could be used for military purposes" is nearly impossible, both because so many technologies are dual-use and because the Chinese market is so large that a radical separation would cause enormous damage even to those who impose it.
The question, then, is how to calibrate these policies. And, frankly, no one has a definitive answer yet."
Where do you draw the line between absolute free trade and the idea that any dependence is a threat to security?
“It's what we study from morning to night.
The analogy I have in mind is financial regulation after 2008. For a while, it seemed like there were only two extremes: either central banks guaranteed everything, or we went back to the idea that markets were perfect and that we could create all sorts of derivatives because crises wouldn't happen anyway.
Through research, experience, and a bit of composure, we've developed an intermediate level of regulation over time. Imperfect, certainly, but capable of limiting certain activities without abolishing finance.
We're still in a much more primitive phase of geoeconomics. We seem to be wavering between "it's all geopolitical bullying" and "let's trade freely, economic dependencies don't exist." We need to reach a more disciplined position.
There are some activities where it is not prudent to rely entirely on China or the United States. There are technologies we must carefully evaluate before exporting because they could strengthen a capability that is then used against us.
But for most of the economy, the answer must continue to be: let the private sector work and keep trading.
Defining where that line lies takes time and research. But it will have a huge impact on our future lives.”
So what we see today is not an episode, but the beginning of a transformation destined to last?
“Yes. It's a rotation that has just begun and will continue.
This doesn't necessarily mean that the overall level of global trade will decline. It could rise or fall. What will change is its nature: which sectors we trade with, with whom, which activities we consider sensitive, and how geopolitical blocs organize themselves. This rotation has already begun. What we are witnessing today is just the beginning.
Finally, there's an institutional problem. Geoeconomics cuts across government, the private sector, foreign policy, and national security. Who should decide? Shouldn't there be a National Security Council?
Even in the United States, the division of responsibilities between the National Security Council, the Treasury, the Department of Commerce, and the State Department often results in highly fragmented policymaking. But there's an even more fundamental issue.
Geoeconomics is a very old and yet very new topic. And it cannot be entirely delegated to experts, as we have done, at least in part, with monetary policy. These decisions have fiscal, industrial, and labor implications. In a democracy, they must therefore be made by the executive branch. And this also means they must be understood by citizens.
If the prime minister and his government decide, that prime minister gets elected. We can't just leave everything to a few bureaucrats. We need to explain to the public what these decisions mean and convince them of the reasons behind them.
If we fail to bring citizens into this discussion, we risk losing them, as has already happened with trade and tariffs. And this, perhaps, is one of the most difficult parts of the job."
saving
webinfo@adnkronos.com (Web Info)