# What if Trump Isn't Crazy?

### World power is shifting. The real question is not understanding Donald Trump: it is whether Europe still has time to adapt.

---

There is a sentence by Major-General Arnaut Moreira that deserves to be taken seriously, even by those who profoundly disagree with Donald Trump: "Trump is not the origin of the chaos. Trump is the consequence of a shift of power in the international system." It is a provocative statement, perhaps excessive in its formulation, but it contains a question that Europe has systematically avoided asking. What if what we interpret as a temporary disturbance of the international order is, in reality, the manifestation of a much deeper transformation of that very order?

This text was born after listening to the podcast "O Domínio da Guerra" (The Domain of War), from Observador, in an episode in which Major-General Arnaut Moreira talks with Maria João Simões. It was that conversation that challenged me to look at the numbers and structure the ideas that follow.

For decades, we grew used to thinking that the world built after the Second World War, and consolidated with the end of the Soviet Union, represented a kind of historical destiny. The United States guaranteed a large part of the West's security. Europe deepened its welfare state. Globalisation progressively integrated economies. China produced cheaply for Western consumers, Russia supplied energy, trade grew and technology brought societies closer together. For Europe, it was an extraordinarily favourable combination. But it was not a law of nature: it was a historical circumstance. And that circumstance is disappearing.

The question, therefore, may no longer be whether Trump is destroying the old order. It is whether that old order was already disappearing before Trump, and whether he, in an instinctive, brutal and frequently erratic way, perceived a change that a considerable part of Europe still refuses to accept. This document seeks to answer that question with data, not with adjectives. The central thesis can be summarised in three linked propositions: first, the redistribution of world economic power is a measurable fact that predates Trump; second, Trump is better understood as a symptom and accelerator of that redistribution than as its cause; third, Europe faces a finite window of time to convert the wealth it still possesses into strategic power, and that window is closing at the pace of its own demography.

---

## I. The world that created the G7 no longer exists

When, in 1975, the leaders of the main industrial democracies began to meet in the format that would give rise to the G7, the group represented much more than a gathering of seven developed economies. The United States, Japan, Germany, France, the United Kingdom, Italy and Canada stood at the centre of a world system whose technology, capital, industry, financial institutions and military power were largely concentrated in the West. The dollar dominated the international monetary system. The main technological and industrial companies belonged to the advanced economies. The great financial centres were located in New York, London, Frankfurt or Tokyo. China was not yet a major economic power, India remained a poor economy, and Indonesia, Brazil, Mexico and Turkey were far from the scale they would come to reach.

The world changed profoundly in just one generation. A rigorous way of observing that transformation is to compare the G7 with the so-called E7: China, India, Indonesia, Brazil, Russia, Mexico and Turkey. Before going further, an essential methodological distinction must be made. The E7 is not an organisation, it is not an alliance, it has no common institutions and no coordinated political strategy. China and India are simultaneously partners and rivals. Russia and Turkey cooperate in some domains and compete in others. Brazil, Mexico and Indonesia have very different international interests. The E7 is, therefore, an instrument of economic analysis, not a geopolitical actor. But as an instrument it is extraordinarily useful, because it shows clearly where world production is moving.

The figures from PwC's long-term study "The World in 2050", published in 2017, tell the story in three moments.

**Table 1. Economic size of the E7 relative to the G7 (aggregate GDP in purchasing power parity)**

| Year | E7 as % of G7 | Historical context |
|---|---|---|
| 1991 | ~35% | End of the Cold War; peak of confidence in the Western model |
| 2016 | ~103% (overtaking accomplished) | China is already the world's largest economy in PPP |
| 2040 (projection) | ~200% | The E7 would reach approximately double the G7 |

*Source: PwC, The World in 2050 (2017). See Chart 1 and methodological notes.*

![Chart 1. E7 relative to the G7 in purchasing power parity](/images/graficos/europa-tem-que-reagir/grafico1-e7-g7.png)

*Chart 1. E7 relative to the G7 in purchasing power parity.*

In 1991, the seven E7 economies represented only about 35% of the economic size of the G7, measured in purchasing power parity. Twenty-five years later, they had already overtaken the G7 on that metric. The same projection estimated that, if growth trends held, the E7 could reach approximately double the economic size of the G7 by around 2040 in purchasing power parity. At market exchange rates, a much more demanding measure for emerging economies, the overtaking would be slower, but PwC still projected an E7 significantly larger than the G7 in 2050.

These projections demand two cautions. The first: we must not turn them into a prophecy. They were produced in 2017, and since then there has been a pandemic, wars, inflation, commercial fragmentation, a Chinese slowdown, technological shifts and a profound transformation of international relations. A projection for 2050 is not a fact about 2050. The second caution is symmetrical: the trend that gave rise to the projections has not disappeared. The most recent IMF data show that emerging and developing economies already account for about 60% of world output measured in purchasing power parity, against about 40% for the advanced economies. The crossing of the two curves occurred around 2007-2008, long before Donald Trump entered politics. The world has already changed.

![Chart 2. Share of world output in purchasing power parity](/images/graficos/europa-tem-que-reagir/grafico2-quota-mundial.png)

*Chart 2. Share of world output in purchasing power parity.*

To give concrete shape to this movement, it is worth recording what the same PwC projection anticipated for the ranking of the largest economies in 2050, in purchasing power parity: China in first place, India in second, the United States in third, Indonesia in fourth. The largest economy of the European Union, Germany, would appear in ninth place. Italy, a founding member of the G7, would fall out of the top twenty. We can, and should, debate every number. But the direction of the movement is unequivocal and converges with every other available source.

---

## II. There are two ways of measuring economic power, and both tell the same story

The G7 versus E7 discussion requires a fundamental methodological caution, frequently ignored in public debate: we cannot use nominal GDP and GDP in purchasing power parity as if they were the same thing. They are not, and the difference has relevant analytical consequences.

Nominal GDP, converted at market exchange rates, measures an economy's international financial capacity. It is particularly relevant for buying technology, companies, resources and equipment produced abroad, for carrying weight in global financial markets and for financing external spending, including military spending. Purchasing power parity, in turn, corrects for price differences between countries and seeks to measure the real quantity of goods and services an economy can produce domestically. An economy can look much larger in PPP simply because wages, housing and domestic services are significantly cheaper. That is why the E7's overtaking of the G7 happened much earlier in PPP than in current dollars.

**Table 2. The two metrics and what each one measures**

| Metric | What it measures best | Strategic relevance | E7 vs G7 situation |
|---|---|---|---|
| GDP at market exchange rates | International purchasing power, global financial weight | Acquisition of technology and external assets, sanctions, financing of alliances | G7 still ahead; convergence projected by 2050 |
| GDP in purchasing power parity | Real domestic productive capacity | Industrial base, mobilisation of resources, sustaining prolonged military effort | E7 overtook the G7 around 2016 |

The two metrics tell different parts of the same story, and both point in a similar direction. The enormous concentration of economic power that characterised the West for much of the twentieth century has ended. That does not mean another bloc has automatically taken its place. It means we have entered a much more distributed system of power, in which real productive capacity, the kind PPP measures best, has shifted decisively away from the historical core of the advanced economies. Anyone in doubt about the strategic relevance of this distinction can observe the war in Ukraine: the capacity to produce munitions, drones and steel in quantity has proved, in a context of prolonged conflict, more decisive than the nominal value of the GDPs involved.

---

## III. The great change is not the impoverishment of the West

It is important to understand precisely what is happening, because the most common diagnostic error consists of confusing loss of relative power with absolute impoverishment. The West is not necessarily getting poorer. The United States can keep growing. Europe can keep growing. European citizens may live better in 2050 than they do today. And, even so, the West can lose a very significant part of its relative power. There is no contradiction whatsoever.

The mechanism is arithmetic. Imagine a European economy that grows 30% over a given period, while the world economy grows 100%. Europe has become richer in absolute terms, but it now represents a substantially smaller share of the world economy. And power is not merely absolute wealth: it is also, and above all, relative scale. It is the capacity to finance research, attract talent, control technologies, project force, set standards, impose sanctions, finance alliances, protect trade routes and build infrastructure. In all these dimensions, what counts is not how much one has, but how much one has compared with the others.

Europe's central problem is therefore perhaps not decline in the classical sense of the word. It is something more subtle and, for that reason, more dangerous: **Europe is losing relative scale in a world where scale has come to matter again.** For decades, this loss could seem barely relevant. A Europe protected militarily by the United States, supplied with relatively cheap energy, integrated into an open world economy and able to import technology produced elsewhere could remain extraordinarily prosperous without controlling any of the factors that sustained that prosperity. But that model depended on external conditions that are no longer guaranteed. Cheap Russian energy disappeared with the invasion of Ukraine. Global commercial openness is fragmenting into blocs. And the American security guarantee has become, for the first time in seventy years, the object of explicit negotiation.

---

## IV. This is where Trump comes in

Perhaps we should, then, look at Donald Trump again. Not to turn him into a strategic genius, not to justify all his decisions, not to interpret every contradictory statement as part of a secret plan. But to ask a more intellectually uncomfortable question: what if part of the diagnosis is correct?

Trump looks at the world economy and sees dependence. He sees essential production chains concentrated outside the United States. He sees advanced semiconductors produced mostly in Asia: the Taiwan Semiconductor Manufacturing Company alone manufactures about 90% of the world's most advanced chips, and the extreme ultraviolet lithography machine indispensable for producing them is made by a single company, the Dutch firm ASML. He sees dependence on critical minerals, whose refining is overwhelmingly concentrated in China. He sees strategic industries that were relocated to competitor countries over three decades of cost optimisation. He sees wealthy European allies who for decades invested significantly less in defence than the United States. He sees China using industrial policy, technology, energy and trade as deliberate instruments of national power.

Trump's conclusion is simple, perhaps too simple: a country that does not control what it depends on progressively ceases to control its own destiny. We can disagree frontally with the solutions. We can consider indiscriminate tariffs economically destructive, and the economic evidence largely supports that criticism. We can consider it deeply counterproductive to antagonise allies. We can reject the idea of turning all international relations into a transactional negotiation. We can consider that unpredictability destroys trust, which is the most valuable asset of any system of alliances. All of that can be true. And it can also be true that the problem Trump identifies is real.

The most revealing fact is that America's strategic transformations already go beyond the person of Donald Trump, and several of them predate his return to power. The CHIPS and Science Act, which mobilised tens of billions of dollars to repatriate semiconductor production, was approved in 2022, during the Biden administration, with bipartisan support. The October 2022 export controls on advanced technology to China, likewise. The Inflation Reduction Act, the largest American industrial policy programme in decades, the same. Technological competition with China has become structural. Supply chains came to be treated as matters of national security. Reindustrialisation returned to the centre of economic policy. Artificial intelligence is treated as essential infrastructure of power. And the pressure on Europeans to take on more responsibility for their own defence runs, with different intensities, through the Obama, Trump I, Biden and Trump II administrations. Trump accelerates, radicalises and dramatises these tendencies. He did not invent them.

---

## V. Perhaps Trump is asking the right question in the worst possible way

This is perhaps the most rigorous way of assessing the phenomenon: Trump can be wrong in the answers and simultaneously right in some of the questions. And the questions are these. Why should a power accept strategic dependence on a potential adversary? Why should it allow essential industrial sectors to disappear, if they may prove indispensable in a crisis? Why should it separate economic policy from national security, when its competitors do not make that separation? Why should an American taxpayer indefinitely finance a disproportionate share of the defence of economically developed allies? Why should it consider that semiconductors, artificial intelligence, energy, space and critical minerals are merely markets, and not components of power?

These questions will not disappear when Trump disappears from politics. And perhaps that is one of the most important conclusions of this whole discussion: Europe has spent a decade responding to the messenger, when what demands a response is the message. A continent that organises its entire strategy around the premise of "waiting for it to pass" is, in practice, betting its future on the domestic politics of another country. That is not a strategy. It is an abdication.

---

## VI. Europe believed it had exited History

The great European project was born precisely from the attempt to overcome the traditional logic of power. After two world wars, Europe sought to replace rivalry with integration, economic borders with a common market, nationalism with institutions, confrontation with negotiation, force with law. It was one of the greatest political achievements of contemporary History, and nothing in this document is intended to diminish it. The error was not doing that. The error was perhaps believing that what worked inside Europe would inevitably come to work across the whole world.

It did not happen. Russia continued to think in terms of territory and strategic depth, as the invasion of Ukraine demonstrated with brutality. China never abandoned the centrality of industrial capacity, which today represents about 30% of world manufacturing output. The United States never truly stopped thinking about technological and military supremacy. India progressively returned to great-power status. Turkey developed an assertive policy of regional influence. Europe, meanwhile, became extraordinarily competent at regulating a world it progressively ceased to control.

It is a harsh sentence, but the data sustain it. Europe regulates artificial intelligence, through the world-pioneering AI Act; the most powerful artificial intelligence companies are mainly American. Europe regulates digital platforms, through the Digital Services Act and the Digital Markets Act; the largest platforms are mostly American. Europe discusses digital sovereignty; more than two thirds of its cloud market is in the hands of three American companies, Amazon, Microsoft and Google. Europe possesses extraordinary scientific research; but the Draghi report documents that only four of the world's fifty largest technology companies are European, and that no European Union company with a market capitalisation above 100 billion euros has been created from scratch in the last five decades, while all six American companies worth more than a trillion dollars were born in that period.

**Table 3. The European regulatory paradox: regulating without producing**

| Domain | European regulatory instrument | Productive reality |
|---|---|---|
| Artificial intelligence | AI Act (2024) | Frontier models dominated by American companies; private AI investment in the US several times higher than in Europe |
| Digital platforms | DSA / DMA | None of the major global platforms is European |
| Cloud | "Sovereign cloud" initiatives, GAIA-X | ~70% of the European market controlled by AWS, Azure and Google Cloud |
| Semiconductors | European Chips Act | European share of world production ~10%; advanced nodes concentrated in Taiwan and South Korea; critical exception: ASML in lithography |
| Personal data | GDPR | Global reference standard, but with no European data giants benefiting from it commercially |

The problem is not regulation itself. A society needs rules, and some European rules have become world references. The problem arises when the capacity to regulate replaces the capacity to produce, when Europe exports norms and imports technology, and when normative power becomes the only power left.

---

## VII. Then comes demography

Perhaps no factor determines the European future as much, and receives as little day-to-day political attention, as demography. The Eurostat numbers are uncomfortably clear.

In 2005, the median age of the European Union's population was 39.6 years. In 2025, it reached 44.9 years. In just twenty years, the European median age increased by more than five years, a speed of ageing without historical precedent in peacetime. In that same year, the European Union counted close to one hundred million people aged 65 or over, more than a fifth of the total population.

But the most important number is not the median age. It is the ratio between the elderly population and the working-age population, because that ratio determines how many workers support each pensioner. In 2004, using the 20-64 age bracket, there were almost four working-age people for every elderly person in the European Union. In 2024, there were fewer than three. Using the conventional statistical definition, 15 to 64, the old-age dependency ratio stood at 34.5% at the start of 2025, and Eurostat's demographic projections point to values of around 50% or higher by mid-century. Put another way: we are moving from a system with almost four workers per elderly person to a system with approximately two.

![Chart 3. The European demographic equation](/images/graficos/europa-tem-que-reagir/grafico3-demografia.png)

*Chart 3. The European demographic equation.*

**Table 4. Fundamental demographic series of the European Union**

| Indicator | 2005 | 2015 | 2025 | ~2050 (projection) |
|---|---|---|---|---|
| Median age (years) | 39.6 | 42.4 | 44.9 | ~48-50 |
| Old-age dependency ratio (65+ / 15-64) | ~25% | ~29% | 34.5% | ~50-57% |
| Working-age people per elderly person (20-64) | ~3.9 (2004) | ~3.4 | <3 (2024) | ~2 |

*Source: Eurostat, population structure and ageing; EUROPOP projections. Projection values indicative.*

The implication is gigantic. A proportionally smaller working population will have to finance, simultaneously, pensions, healthcare, long-term care, education, infrastructure, the energy transition, public debt service, technological investment and, now, also much higher defence spending. It is probably here that we find the true equation of the European future, and it is worth formulating it precisely: **how to finance the welfare state of the twentieth century, the defence needed in the twenty-first century and the technological revolution of the twenty-first century with a proportionally older population?**

---

## VIII. The European triangle: three simultaneous demands

Europe faces three demands which, in isolation, would each be a generational challenge, and which present themselves at the same time.

The first is preserving the welfare state. Ageing mechanically increases the pressure on pensions, health and long-term care, and does so precisely when the contributory base is shrinking in relative terms. The second is recovering strategic capacity: defence, energy, critical infrastructure, raw materials and security, domains Europe subcontracted, explicitly or implicitly, for decades. The third is financing the next technological revolution: artificial intelligence, semiconductors, computing, robotics, biotechnology and new forms of energy, the sectors that will determine the economic hierarchy of the coming decades.

Any one of these challenges would require enormous resources. The three simultaneously demand a profound transformation of the European economy, because the alternative, indefinitely increasing spending in all areas while keeping the same growth, the same productivity and the same economic organisation, is arithmetically impossible. The mathematics will eventually impose itself, with or without political decision. The only real choice is between deliberate adaptation and forced adaptation.

---

## IX. The Draghi report is a warning, not a prosperity programme

When Mario Draghi presented, in September 2024, his report on European competitiveness, the diagnosis was extraordinarily clear and convergent with everything said above. Europe faces a productivity problem: the GDP per capita gap with the United States has widened to about 30%, and almost all of that difference is explained by productivity, not hours worked. An investment problem. A scale problem. An energy problem: European companies pay two to three times more for electricity than American ones, and four to five times more for natural gas. A technology problem. And a problem of strategic dependence in practically all critical value chains.

The report estimated additional investment needs in the order of 750 to 800 billion euros per year, approximately 4.5% of the European Union's GDP, to put the European economy on a trajectory compatible with the technological, energy and strategic challenges it faces. It is a number difficult to imagine, and so it helps to give it historical scale: the Marshall Plan, the canonical reference for a great reconstruction effort, represented between 1% and 2% of the GDP of the beneficiary countries. What Draghi proposes is a relative effort two to three times greater than the Marshall Plan. And it is not 800 billion once: it is an annual order of magnitude, sustained over a decade or more.

![Chart 5. The scale of the investment effort](/images/graficos/europa-tem-que-reagir/grafico5-investimento.png)

*Chart 5. The scale of the investment effort.*

But the real European problem is perhaps not even a lack of money. Europe possesses enormous wealth and one of the highest savings rates in the world: European households save about 1.4 trillion euros annually, and the Letta report estimated that about 300 billion euros of European savings are channelled every year out of the Union, mostly to American markets, where they find depth, liquidity and returns that European markets do not offer. The problem is the difficulty of transforming that capital into productive investment at continental scale.

The reason for that difficulty has a name: fragmentation. The Union has 27 capital markets still insufficiently integrated, 27 tax systems, 27 industrial priorities, 27 governments protecting national champions and 27 military procurement systems. And then we are surprised that American companies, born into a domestic market of continental dimension and a deeply integrated financial system, reach scale faster than European ones. The conclusion is uncomfortable but inescapable: Europe possesses scale, but uses it badly.

---

## X. A power of 450 million behaving like 27 medium-sized countries

This is perhaps the fundamental European paradox. Individually, none of the main European countries possesses the scale of the United States or China, and in the future will not have the scale of India either. Germany, France, Italy, Spain and Poland will remain relevant countries, but none will manage, in isolation, to competitively finance all the infrastructure needed to lead simultaneously in artificial intelligence, space, defence, energy, semiconductors, advanced computing and biotechnology. No economy of 50 or 80 million inhabitants can sustain that portfolio of ambitions alone in a world where competitors have 330 million, 1,400 million and unified capital markets.

Collectively, Europe has scale: 450 million inhabitants, the second largest consumer market in the world, a first-rate scientific base. Individually, it is condemned to lose it. The question of the European future perhaps reduces, therefore, to a simple question: will Europe manage to transform economic size into political and strategic power before the relative loss of weight makes that transformation much more difficult? Because there is a cruel temporal dynamic here: each year of fragmentation reduces Europe's relative weight, and each reduction of relative weight diminishes the capacity to negotiate, attract and finance its own integration. The window is not infinite.

---

## XI. And now we have to pay for defence

For decades, much of Europe lived a historically exceptional situation: it built some of the most generous social systems in the world without fully assuming the strategic cost of its own protection. The United States bore a very significant share of NATO's military capacity, still today in the order of two thirds of the Alliance's total spending. Europe benefited and was able to invest in other priorities. It was, in financial terms, the largest strategic subsidy in modern history. That period has ended.

At the NATO summit held in The Hague, in June 2025, the allies committed to move by 2035 towards investment equivalent to 5% of GDP in defence and security, broken down into at least 3.5% for core military needs and up to 1.5% for areas related to resilience, critical infrastructure, innovation and defence industrial capacity. To measure the scale of the transformation, recall the starting point: in 2014, when the Wales summit set the 2% target, European allies spent on average about 1.5% of GDP; in 2024, they were finally around 2%.

![Chart 4. From subsidised protection to rearmament](/images/graficos/europa-tem-que-reagir/grafico4-defesa.png)

*Chart 4. From subsidised protection to rearmament.*

The arithmetic is implacable. An economy that goes from spending 1.5% to 3.5% of GDP on defence transfers two additional percentage points of its output to that purpose every year. In an economy of one trillion euros, roughly the scale of the Spanish or Dutch economy, that represents an additional 20 billion euros. Not once: every year. At the scale of the European Union, with a GDP in the order of 17 to 18 trillion euros, each additional percentage point of defence spending represents about 170 to 180 billion euros annually. Defence has therefore returned to the great European contest for resources, and it returns precisely at the moment when needs are increasing with ageing, technology, energy and the industrial transition. The triangle described in section VIII is not a metaphor: it is a budget constraint.

---

## XII. Spending more does not necessarily mean having more power

Here lies one of the most important, and least discussed, risks of European rearmament: Europe can brutally increase its military spending and remain strategically dependent. The recent data are a warning. According to the Draghi report, between mid-2022 and mid-2023, 78% of European defence procurement spending was made outside the European Union, and 63% with American suppliers. If Europe essentially buys equipment produced in the United States, it will have more immediate military capacity, but it will reinforce the technological dependence that rearmament was supposed to reduce.

Fragmentation aggravates the problem. If 27 countries keep buying different systems, the current absurd duplication persists: European armed forces operate about a dozen different models of battle tank, while the United States operates essentially one. Logistics lines, maintenance costs, training programmes and operational incompatibilities multiply. And if each government uses the new defence budgets mainly to protect its national champions, the opportunity to build European defence companies with global scale will be wasted.

The right question is therefore not just "how much are we going to spend?". It is: **what power do we want to build with the money we are going to spend?** Because defence can be much more than military expenditure. It can be industrial policy, research, artificial intelligence, robotics, drones, satellites, space, cybersecurity, new materials, energy, communications and advanced computing. Recent American history is instructive: a large part of the United States' technological revolution of recent decades, from the internet to GPS, from semiconductors to artificial intelligence, benefited directly or indirectly from the state's strategic investment, frequently through the defence budget and agencies such as DARPA. Europe stands before a fork in the road: it can use rearmament to rebuild its own industrial and technological capacity, or it can simply become the largest external client of the American military industry. The economic and strategic consequences of these two choices are radically different, and the decision is being taken now, contract by contract.

---

## XIII. Artificial intelligence may be the answer to demography

There is one variable capable of partially altering the terms of the European triangle: productivity. A society with fewer workers is not inevitably condemned to produce less; it needs each worker to produce more value. This is where artificial intelligence takes on, for Europe, a qualitatively different importance from the one it has for the United States. For the American economy, with more favourable demography and sustained immigration, AI is an extraordinary growth opportunity. For Europe, it may become a demographic necessity.

The applications are concrete and transversal: automating administrative processes in a continent whose bureaucracy is notoriously heavy; increasing industrial productivity in sectors where Europe still holds advantages; accelerating scientific research; supporting health systems pressured by ageing; transforming public services; using robotics in sectors with structural labour shortages, from construction to elderly care; and expanding the capacity of small and medium-sized enterprises, which form the backbone of the European economic fabric. AI can allow a smaller working population to sustain a larger economy. It does not solve every problem, but it profoundly changes the equation: it is the only variable of the triangle that acts simultaneously on all three vertices, financing the welfare state through productivity, reinforcing strategic capacity and constituting, itself, the technological revolution in question.

Hence the inescapable conclusion: Europe cannot limit itself to being the continent that regulates other people's artificial intelligence. It needs to own it, develop it, deploy it, use it and scale it. A continent that faces the European demographic deficit and treats the technology that can compensate for it primarily as a risk to be contained, rather than a capacity to be built, is committing a strategic error of the first order.

---

## XIV. And then there is immigration

No serious discussion of the European demographic future can avoid this topic, however politically inflammable it may be. The arithmetic is simple: Europe needs workers. But immigration is not simply an economic variable to be adjusted in a model. It is also housing, schools, culture, identity, integration, security, social cohesion and administrative capacity. The idea that importing population is enough to solve ageing is too simple, and recent European experience demonstrates the political costs of treating it as such.

But the opposite idea is equally unsustainable. A Europe that ages rapidly, closes its borders completely and keeps fertility rates in the order of 1.4 to 1.5 children per woman, far below the replacement level of 2.1, will hardly manage to preserve simultaneously economic growth, pensions, healthcare and public services without productivity increases of a magnitude no developed economy has ever achieved in a sustained way.

European politics will therefore have to find a balance it has not yet found: attracting the people it economically needs and building societies capable of integrating them socially. The two terms are inseparable, and the failure of either feeds the failure of the other. Failure at integration will feed political movements seeking to close the borders; complete closure will worsen the demographic problems, which will degrade public services and feed new discontent. It is one of the most difficult political dilemmas of the coming decades, and one of those in which the absence of a strategy is, in itself, a strategy: the worst one.

---

## XV. Five possible Europes in 2050

There is no inevitable future. There are choices, and it is useful to make them explicit in the form of scenarios. None of these five futures is a prediction; each is the disciplined extrapolation of a set of decisions, or non-decisions, taken over the next ten years.

**Scenario 1: Museum Europe.** Europe remains rich, beautiful, safe and socially sophisticated. It is excellent to live in, excellent to visit, excellent to study in. But it depends on others for almost everything that determines power: American defence, American cloud, American artificial intelligence, Asian semiconductors, external raw materials, imported energy. It would not necessarily be poor; it would be dependent. This is perhaps the most dangerous scenario, precisely because it can remain comfortable for a long time, indefinitely postponing the moment of recognition.

**Scenario 2: Fortress Europe.** Faced with insecurity, immigration and economic competition, each country turns inward: more nationalism, more protectionism, more borders, more national industrial policy, more national military spending. The problem is structural: no European country possesses, alone, the scale of the great global powers. A Europe of 27 fortresses may feel more sovereign and become collectively less sovereign. It is the scenario in which the emotion of sovereignty destroys the substance of sovereignty.

**Scenario 3: American Europe.** Europe accepts permanent strategic dependence on the United States: it buys American technology, American weapons, American cloud, American artificial intelligence models, and keeps NATO as the fundamental security guarantee. It can be a rational choice, and for seventy years it worked. But it means accepting that a relevant part of European autonomy exists within the limits, and the goodwill, of the transatlantic relationship. Trump made the risk of this option visible: dependence is comfortable as long as whoever protects us wants to keep doing so under the conditions we expect. It is no longer possible to assume they will.

**Scenario 4: Irrelevant Europe.** There is no collapse; there is erosion. Low growth, insufficient productivity, high debt, ageing, investment flight, technology companies leaving to grow in other markets, persistent fragmentation, incapacity to decide. Nobody announces the day Europe stops being a great power. There simply comes a moment when the fundamental decisions are taken in Washington, Beijing and New Delhi, and Brussels reacts. Irrelevance is not decreed: it accumulates.

**Scenario 5: Europe as a power.** There is another possibility, and its components are known because they are described, in detail, in the Draghi and Letta reports: a true integration of capital markets, a continental energy strategy, a European defence industry with coordinated military procurement, common artificial intelligence infrastructure, its own capacity in critical technologies, scientific investment at continental scale, a truly single market for digital companies, a common policy for raw materials and a migration strategy that is simultaneously economic and social. It does not necessarily mean immediately creating the United States of Europe. It means recognising that there are domains in which isolated national sovereignty has ceased to produce real sovereignty, and acting accordingly.

**Table 5. Scenario synthesis**

| Scenario | Main driver | Wealth in 2050 | Power in 2050 | Reversibility |
|---|---|---|---|---|
| Museum Europe | Comfortable inertia | High | Low | Decreasing |
| Fortress Europe | Defensive nationalism | Medium/low | Fragmented | Difficult |
| American Europe | Strategic subcontracting | High | Conditioned | Dependent on third parties |
| Irrelevant Europe | Unmanaged erosion | Medium | Residual | Very difficult |
| Europe as a power | Selective integration and scale | High | High | Self-sustaining |

---

## XVI. Trump's problem may be the method

This is where we return to the starting point. Perhaps Trump is right when he says Europe needs to pay more for its own security. He may be right when he states that industrial dependence is dangerous. He may be right when he treats technology, energy and industry as components of national power. He may be right when he considers that the old globalisation created strategic vulnerabilities. None of this means he is right in what he does with that diagnosis.

Antagonising allies may destroy one of the greatest strategic advantages of the United States. The Western democracies, when combined, still represent an extraordinary concentration of capital, technology, knowledge and military power: more than half of world military spending, the majority of frontier scientific research, the main financial centres. China has partners; the United States has allies. It is not the same thing, and the difference is worth more than any tariff will ever yield. Tariffs can protect certain industries and, simultaneously, raise costs for many others, functioning as a diffuse tax on one's own economy. Unpredictability can create momentary negotiating capacity; it also destroys trust, which is the accumulated capital of eight decades of American-led order.

Trump may, therefore, be right about part of the disease and profoundly wrong about the medicine. And from this follows the essential conclusion for this side of the Atlantic: Europe does not need to become Trumpist to recognise the diagnosis. It only needs to stop confusing criticism of the messenger with refutation of the message.

---

## XVII. Perhaps the European error was believing in permanence

For decades, Europe believed that peace was permanent. That cheap energy would remain available. That the United States would always keep paying for security. That industry could be relocated to other countries without strategic consequences. That technology would always be accessible on the market, at market price. That interdependence would prevent conflicts. That regulating a technology was almost as important as mastering it. That it was possible to age, grow little, invest insufficiently and indefinitely maintain one of the most generous social models in History.

Each of these beliefs was understandable in its context, and several were rational for a long time. Perhaps the great European strategic error was not each of these decisions in isolation. Perhaps it was believing that all of this could be done simultaneously and forever. History, as someone wrote, does not end: it only pauses. The European pause lasted three decades. It ended somewhere between the annexation of Crimea, the pandemic, the invasion of Ukraine and the return of the word "power" to the vocabulary of Brussels.

---

## XVIII. 1975-2050: four acts of a historic shift

The transformation described throughout this document can be condensed into a chronology of four acts.

**1975: the West at the centre.** The great industrial democracies dominate the economy, technology, capital and military power. China is not yet a global economic power. Europe grows protected by the American alliance, and the G7 is born as the institutional mirror of that hegemony.

**1991: the illusion of definitive victory.** The Soviet Union disappears. The E7 represents only about 35% of the size of the G7 in purchasing power parity. Liberal democracy and the market economy seem to have won the century's great ideological contest, and that appearance of final victory becomes, paradoxically, the cradle of complacency.

**2016-2026: power returns.** The E7 overtakes the G7 in PPP. Emerging and developing economies come to represent the majority of world output on that metric. China and the United States enter open technological competition. Territorial war returns to Europe. Energy, semiconductors and artificial intelligence become matters of national security. NATO drastically redefines defence investment targets. Europe rediscovers the word "power".

**2050: the choice.** Europe will have aged: dependency ratios in the order of 50%, about two workers per elderly person. It will probably represent a smaller share of the world economy. It will have to finance much more defence and will need to produce much more with a proportionally smaller working population. The only open question is whether it will arrive in 2050 more integrated and strategically autonomous, or richer in assets and poorer in power.

---

## XIX. What if Trump saw it first?

Perhaps Donald Trump is profoundly wrong in many answers. Perhaps he is wrong in the way he treats allies. Perhaps he is wrong in the use of tariffs. Perhaps he frequently confuses unpredictability with strategy, and perhaps part of what looks like calculation is simply improvisation. But there is a possibility Europe cannot afford to ignore: the possibility that Trump is right about the fundamental question. The world that allowed the West to live for decades as the economic, technological and strategic centre of the planet is disappearing.

It is worth being rigorous about what this means and what it does not mean. It does not mean the end of the West. It does not mean the inevitable collapse of Europe. It does not mean that China will come to dominate the world: projections are not destinies, China is also ageing, and at an even faster pace than Europe; Russia has deep economic and demographic limitations; India faces enormous internal challenges; and the E7, remember, is not an alliance. The world may become not less Western because another bloc replaced it, but simply much more multipolar. And it is precisely in that world, more distributed, more competitive and less indulgent, that Europe will have to decide what it intends to be.

The answers do not require Europe to renounce what it is. Europe does not need to become Trumpist; it needs to become strategic. It does not need to abandon the welfare state; it needs to build the productivity capable of financing it. It does not need to abandon regulation; it needs to also create again what it regulates. It does not need to abandon the American alliance; it needs to be strong enough for that alliance to be a choice, not an absolute dependence. It does not need to abandon its values to recover power; it needs to understand that values only truly influence History when there is the capacity to defend them.

Perhaps, for that reason, the great question of the next twenty years is not whether Europe will remain a good place to live. It probably will. The question is another: **will Europe still have enough power to decide the conditions in which it lives?**

Trump will pass. The redistribution of world power will remain. Europe's real choice is not between Donald Trump and the world that existed before Donald Trump, because that world no longer exists. The choice is between adapting while it still possesses enough wealth, knowledge, institutions and scale to determine its future, or continuing to carefully administer the present until it discovers it no longer controls tomorrow.

For decades, we asked how much it would cost Europe to truly become a power. Perhaps it is time to ask the inverse question: **how much will it cost us not to?**

---

## Methodological notes and sources

**On the E7 concept.** The E7 (China, India, Indonesia, Brazil, Russia, Mexico, Turkey) is an analytical construction popularised by PwC in the "The World in 2050" series of studies. It does not correspond to any organisation, alliance or political bloc. All E7/G7 comparisons in this document are based on the February 2017 report, produced before the pandemic, the war in Ukraine and the current commercial fragmentation; the values for 2040 and 2050 should be read as conditional projections, not as predictions.

**On the GDP metrics.** Comparisons in purchasing power parity (PPP) use the conversion factors of the International Comparison Program, reflected in the IMF (World Economic Outlook) and World Bank databases. The shares of world output of emerging and developing economies (in the order of 60% in PPP in the most recent IMF data) refer to the IMF classification's "Emerging Market and Developing Economies" aggregate, broader than the E7. The values in Chart 2 are approximate and rounded; for publication use, direct extraction of the series from the most recent edition of the World Economic Outlook is recommended.

**On demography.** EU median age (39.6 years in 2005; 44.9 in 2025), population aged 65 or over (close to one hundred million) and old-age dependency ratio (34.5% at the start of 2025, definition 65+/15-64) come from Eurostat (population structure and ageing indicators). The ratio of almost four working-age people (20-64) per elderly person in 2004, and fewer than three in 2024, comes from the same series. The projections for 2050 are based on Eurostat's EUROPOP exercises and should be treated as ranges, not point values.

**On competitiveness and investment.** The additional investment needs of 750 to 800 billion euros per year (about 4.5% of EU GDP), the productivity gap with the US, the energy price differentials, the statistic on the world's largest technology companies and the defence procurement data (78% extra-EU, 63% from US suppliers, between mid-2022 and mid-2023) come from M. Draghi, "The Future of European Competitiveness" (September 2024). The estimate of about 300 billion euros per year of European savings channelled out of the Union comes from E. Letta, "Much More Than a Market" (April 2024). The comparison with the Marshall Plan (1% to 2% of beneficiaries' GDP) is a conventional historical order of magnitude, used by the Draghi report itself.

**On defence.** The targets of 5% of GDP by 2035 (3.5% in core military capabilities and up to 1.5% in resilience and broader security) come from the declaration of the NATO Summit in The Hague (June 2025). The historical spending averages of the European allies (about 1.5% in 2014; about 2% in 2024) come from the annual publications "Defence Expenditure of NATO Countries". The reference to the diversity of European weapons systems (about a dozen battle tank models versus one in the US) is cited in the Draghi report.

**On technology.** The concentration of advanced semiconductor production in TSMC (in the order of 90% of the most advanced nodes), ASML's monopoly in EUV lithography and the share of the three large American companies in the European cloud market (in the order of 70%) correspond to widely published market estimates (SIA, TrendForce, Synergy Research) cited in official European documents; they should be verified at source at the date of publication.

**Status of the document.** This text is an analytical working basis intended for reflection and discussion. The numbers were selected for their robustness and traceability, but any editorial use should confirm the values in the primary sources indicated, in particular those depending on annually updated editions (IMF, Eurostat, NATO).

**On the starting point.** The reflection that gave rise to this text was sparked by the podcast "O Domínio da Guerra" (The Domain of War), from Observador, with Major-General Arnaut Moreira in conversation with Maria João Simões. Available at: https://observador.pt/programas/o-dominio-da-guerra/