€264 billion per year. That’s the sum European companies inject into the American economy through their purchases of software and cloud services. A recurring cost, invisible at first glance, but heavy with macroeconomic and strategic impacts. The Asterès report takes the measure: Europe buys, the United States captures the value and employment. And if part of this value were reinternalized, the study quantifies tangible gains for the Union. The debate is not new. What’s new here is the quantification effort, based on field interviews and an economic impact model.
An Economic Survey from the Field
Asterès confronted a classic obstacle: statistics on digital services trade remain incomplete, “aggregated” and blurred by accounting realities (tax locations, profit repatriations). To circumvent the obstacle, the firm conducted interviews with six CIOs from major European companies. Objective: estimate credible orders of magnitude then “calibrate” them to market size. The economic footprint was then modeled using the Asterès Impact Model (MIA). Pragmatic approach: start from CIO practice, then work up to macro effects.
The scope is strict. We’re talking exclusively about cloud-software purchases by European companies from American players, for services produced in the United States. When they are produced in Europe by American subsidiaries, the analysis shifts to FDI revenue side. No hardware, no general outsourcing, no public or household purchases: the core subject is the economic foundation of cloud platforms and publishers.
The Blind Spot in Statistics, and the “Ireland Effect”
WTO or IMF series struggle to isolate the “cloud-software” subcategory. Above all, Ireland clouds the picture: an accounting hub, it records a large share of service exports to the entire world, including the United States. Result: “bilateral” Europe–USA flows appear underestimated if we rely solely on service customs data. Hence the value of triangulation: CIO interviews, market shares, and spending orders of magnitude.
The study goes further: it suggests that the majority of international cloud-software flows transit through profit repatriations (FDI) rather than through “exports” in the strict sense. The differential between measured direct imports and spending actually addressed to American players feeds this hypothesis.
The Figure That Changes the Scale: €264 Billion Per Year
At the end of this reconstruction, Asterès estimates at €264 billion per year the cloud-software purchases by EU companies that benefit the American economy. In these expenditures, the share of American players stands at approximately 83% of the European cloud-software market; and 80% of the associated added value would be created on American soil. The order of magnitude speaks for itself: the annual amount is of the same order as the EU’s “energy bill” in 2024 (€360 billion).
The effect is not marginal: it weighs on the European balance of payments and externalizes innovation spillovers, skilled jobs, and taxation. The finding echoes recurring warnings about digital dependency: no European company appears among the top eight global cloud players, the study notes.
The American Footprint: 1.9 Million Jobs Supported
On the US side, the footprint is clear. According to the MIA model, these European purchases directly support approximately $285 billion in revenue and $186 billion in added value in the United States. Including indirect and induced effects, we reach $323 billion in added value (i.e., 1.1% of US GDP) and 1.9 million jobs (approximately 1.2% of total employment). A solid production base, consolidated by European spending.
Put another way: each euro spent in Europe on these segments diffuses a substantial portion of value into American chains, with local tax spillovers (estimated at $89 billion in total, direct + indirect/induced). Digital dependencies are therefore not a theoretical concept; they are observed in national accounts.
What Would Happen If Europe Repatriated a Share of the Market?
Asterès explores three scenarios for reorienting purchases currently addressed to the American economy:
- 5% reoriented: 178,000 additional jobs in Europe.
- 10% by 2030: 331,000 jobs.
- 15% by 2035: 463,000 jobs, €37 billion in added value, and €16 billion in public revenue.
Modest orders of magnitude as % of GDP, but politically significant, as they are concentrated on skilled jobs and critical value chains.
These figures don’t say that a “sovereign cloud” can be decreed. They indicate that, if the EU gains market share, economic multipliers do indeed exist. Provided we maintain the course on technological offering, cost competitiveness, regulatory compliance… and the long term.
Rising Prices, Shifting Balance
Another weak signal highlighted: price dynamics. According to the CIOs interviewed, cloud-software prices are rising by approximately 10% per year, against a backdrop of lock-in, bundling, and low substitutability once the environment is deployed. Mechanical effect: if this trend continues, American cloud-software service exports to the EU would inflate by approximately €421 billion over ten years, significantly improving the American current account balance. European mirror: a partial rebalancing (15% of purchases reoriented) would generate a €100 billion gain in current account balance over ten years for the EU.
Here again, the study remains cautious: it isolates the price effect and assumes, all else being equal, stable trade openness. In practice, geopolitical, energy, or regulatory tensions can accelerate or slow these trajectories.
Productivity: A Potential of +1.2 pts… and an Open Debate
One of the report’s contributions concerns productivity. If the European digital sector (in the sense of IT & other information services) achieved the relative productivity observed in the United States, the EU would register a total gain of approximately 1.2% in productivity. Why? Because in the United States, digital is significantly more productive than the economy average, whereas in Europe its productivity is close to average. Aligning the European profile with the American profile would therefore raise overall productivity, even though the sector only accounts for slightly less than 2% of employment.
But the study qualifies, echoing the Draghi report. The productivity slowdown in Europe since the 1990s is not explained solely by digital, and international comparisons involve measurement biases (sectoral level, PPP, ILO series, etc.). Several studies also suggest a recent stagnation in the overall US/EU differential. In short, the digital effect is likely, but the precise estimate remains fragile; it calls for additional research.
Blind Spots to Address
Four areas emerge, useful for public decision-makers as well as investment committees:
- Better measure cloud-software service exchanges. Without reliable series, we debate blindly. “Trade” and “repatriated profits” flows must be tracked together, with harmonized sectoral granularities (Eurostat, WTO, BEA, etc.).
- Anticipate geopolitical dependency risks. The report highlights a potential lever: reducing the supply of critical services is a coercion weapon. The signal is clear for strategic sectors (health, energy, transport, defense).
- Monitor the price effect on user companies’ financial health. Double-digit increases can erode margins, capex, and transformation trajectories, particularly in finance and digital-intensive sectors.
- Link productivity and competitiveness without confusing the two. A productivity gain does not automatically translate into a price-competitiveness gain if wages and non-wage costs evolve in parallel. Competitive advantage depends on a mix of factors (service quality, image, compliance, security, environmental footprint).
What Credible Action Agenda for Europe?
The Asterès study does not prescribe public policies. It provides orders of magnitude that help prioritize levers. Several paths emerge when reasoning in political economy, sector by sector.
- Accelerate European cloud upscaling. The objective is not autarky, but counterweight. This requires heavy investments (datacenter capex, networks, interconnections), open architectures (interoperability, reversibility), and trust standards (security, sectoral compliance, sobriety). Without application ecosystems (vertical SaaS) and solid distribution chains, infrastructure alone is not enough. The 5–15% reorientation scenarios show that each market share point counts, but requires a credible offering.
- Put public procurement in strategic alignment. Group purchasing, technical reversibility clauses, penalties for unjustified price increases, interoperability and data portability requirements: so many tools to reduce lock-in and create a contestable playing field. The CIOs interviewed point to trending upward prices and difficulty of substitution after deployment; purchasing power can correct this bias.
- Target priority verticals. The battle is won sector by sector. Health, finance, industry, energy, mobility: supporting European SaaS that addresses specific compliance and sovereignty needs (traceability, auditability, localization, certification) is a shortcut to product-market fit. The pull effect on infrastructure will follow.
- Enable cloud mobility. Generalize multicloud architecture frameworks, workload portability, and standard management interfaces (observability, security, costs). Anything that reduces switching costs makes competition effective, even in the presence of dominant players.
- Establish a European observatory of prices and practices. Transparency on price formation, bundling, and licensing models would create an informational counterpower. It would also feed foresight (price effect on user margins and capex, service productivity, sectoral exposure).
- Connect skills and market. A talent strategy (DevOps, SRE, FinOps, cybersecurity, MLOps) conditions competitiveness as much as material investment. Without a skills pool, no scalability or time-to-market.
What the Study Tells Us… and What It Doesn’t
The strength of the Asterès report lies in its pedagogy: it makes visible intangible flows that weigh on the real economy. The quantification—€264 billion per year—provides a basis for discussion. US multipliers and European scenarios illuminate the magnitude of potential gains. The limitations are acknowledged: statistical gaps, dependence on interviews, structural difficulties in isolating the “cloud-software” category, sensitivity of price assumptions. Asterès does not claim to settle the matter. It prioritizes questions to address: measurement, geopolitics, price effect, productivity-competitiveness.
For decision-makers, the issue is not binary—”all US” or “all EU.” It’s portfolio: controlled diversification, interoperability, local value chains where relevant, and rebound capacity if the environment tightens. The right metric is not self-sufficiency, but the ability to steer one’s digital spending.
References
Asterès (April 2025) – Technological dependency on American software & cloud services: an estimate of economic consequences in Europe, study for Cigref.
Draghi, M. (2024) – The Future of European Competitiveness (part A).
