Long read · economics · Draft v1 · October 2026 · sources and items to verify at the end.

1 Introduction

The United States will keep its growth lead over Europe through 2031: the IMF expects 2.3% real growth for the US in 2026 and 2027 against 1.1% and 1.2% for the euro area (IMF, 2026). The usual explanation, that Europe is uniformly over-taxed and over-regulated, is wrong in one respect. Europe is two models sharing a currency, and the average of the two describes neither.

The first model, Austria, Germany, France, Belgium, the Netherlands and the Nordics, combines high income with high charges on labour, expensive energy and dense administration. An Austrian owner distributing 1 million euros of profit keeps about 558,000 euros, and the top income-tax rate of 55% runs to the end of 2029 (Tax Foundation, 2026). The second model, Romania, Bulgaria, Hungary, Poland and the Baltics, runs flat taxes of 10% to 16%, labour at 12 to 20 euros an hour and output per head at 68% to 90% of the EU average (Eurostat, 2026a). Averaged, the EU shows a corporate rate of 21.6% against 25.6% in the US; no firm pays it. The article therefore compares pairs: Austria and Germany with Massachusetts, New York and California on tax structure; Romania and Bulgaria with Mississippi and West Virginia on income; Estonia and Ireland with Texas and Florida as low-tax locations.

The argument runs in four parts. Part I measures the gap and finds that it is a gap in dynamics, not in levels. Part II explains why the European core is expensive: a state that spends more than half of output, a tax stack that falls on labour rather than capital, energy at three to four times the Texan price, and rules that are honest but slow. Part III follows the consequence: thin risk capital, a cautious investment climate, a technology market in which US firms took about 75% of global AI venture capital in 2025 against 6% for the EU-27 (OECD, 2026a), and a talent flow that has begun to reverse. Part IV turns these forces into scenarios to 2031.

Part I: The Gap

2 Size, Income and Growth

In current dollars the US economy is about 1.4 times the EU’s, in purchasing-power parities about 1.06 times: IMF projections for 2026 put nominal GDP at 32.4 trillion dollars for the US and 23.0 trillion for the EU (IMF via StatisticsTimes, 2026). Per head the US leads by about 85% nominally and 39% at purchasing power; Eurostat puts EU GDP per head at 41,566 euros in purchasing power standards (PPS) for 2025 (Eurostat, 2026a). Two caveats: Irish GDP per head stands at 237% of the EU average while its consumption per head is at the average, since multinational profit-booking inflates both figures (Eurostat, 2026b), and US GDP counts health spending that produces no better outcomes.

Table 1: GDP per head in PPS, real growth, deficit and debt 2025 (EU = 41,566 euros)

Tier Country GDP per head (PPS) Growth 2025 Deficit 2025 Debt 2025
High income, high burden Netherlands 55,152 euros 1.6% −1.6% 44.4%
Denmark 53,941 euros 3.5% +2.9% 27.9%
Austria 49,022 euros 0.8% −4.2% 81.5%
Germany 48,622 euros 0.2% −2.7% 63.5%
Belgium 47,676 euros 1.0% −5.2% 107.9%
Sweden 46,344 euros 1.4% −1.3% 35.1%
France 40,716 euros 0.8% −5.1% 115.6%
Italy 39,949 euros 0.5% −3.1% 137.1%
Low tax, high income Luxembourg 99,621 euros 0.6% −2.0% 26.5%
Ireland 93,360 euros 8.0% +1.8% 32.9%
Low tax, lower income Czechia 38,557 euros 2.6% −2.1% 44.3%
Lithuania 36,174 euros 2.9% −1.8% 39.5%
Poland 33,816 euros 3.6% −7.3% 59.7%
Estonia 33,279 euros 1.3% −2.0% 24.1%
Romania 32,353 euros 0.7% −7.9% 59.3%
Hungary 31,719 euros 0.5% −4.7% 74.6%
Bulgaria 28,332 euros 3.1% −3.5% 29.9%

(Source: Eurostat, 2026a, 2026c; Austria’s 2025 growth is 1.0% in the IMF Article IV report, and WIFO expects 0.9% for 2026)

US states spread as widely as EU members: BEA nominal GDP per head for 2025 is about 123,000 dollars in New York, 114,000 in Massachusetts, 109,000 in California, 90,000 nationally, 62,000 in West Virginia and 55,000 in Mississippi (BEA, 2026). On World Bank purchasing-power data Austria stood at 63,970 dollars in 2025, Romania at about 41,000 and Bulgaria at about 35,000, while the IMF’s October 2026 outlook gives Austria about 76,000; Switzerland reaches about 85,700. Three pairs follow: Romania and Bulgaria match Mississippi and West Virginia on income; Austria and Germany match Massachusetts and California on tax and cost structure but sit closer to Kentucky or Alabama on income; only Luxembourg, Ireland and Switzerland match the strongest US states. An Austrian entrepreneur thus carries a California-grade burden on a below-average American income base. Section 3 asks how much of that gap is productivity rather than price level.

3 Productivity: A Stable Level, a Widening Dynamic

The level gap has been stable for 25 years; the dynamics gap has not. US output per hour has risen 9.7% since 2019, EU output 2.4%, and in 2024 the two stood at 116 and 72 dollars per hour in current purchasing-power parities (World Employment Confederation, 2025). The OECD reports 84.1 dollars for the US and 135.7 for Ireland in constant 2020 parities, noting that a gross-national-income measure would place Ireland near the other high-income European economies (OECD, 2026b).

Three readings of the gap exist. ITIF attributes 72% of the US lead of 34% at purchasing power to lower productivity per hour and 28% to fewer hours (ITIF, 2025). The Banque de France found the hours gap narrowed from about 20% in 2000 to 10% in 2023, with shorter working time still the largest factor in Germany and France (Banque de France, 2025). Bruegel and Krugman hold the level ratio constant over 25 years, with the eastern members rising from 32% to 55% of US income per head between 1995 and 2022 (Bruegel, 2025). A 2026 Richmond Fed brief reconciles them: levels stable, growth lag real (Richmond Fed, 2026). This article follows that reading, since the lag concentrates in software and AI [PLACEHOLDER: OECD GDP per hour worked 2024 for AT, DE, FR, NL, SE, PL, RO].

Research intensity in the core is high but uneven: only Sweden (3.6%), Belgium (3.4%), Austria (3.3%), Finland (3.2%), Germany (3.1%) and Denmark (3.0%) spend 3% or more of GDP on R&D, against 2.24% for the EU in 2024 and about 3.5% for the US on older OECD data (Eurostat, 2025) [PLACEHOLDER: OECD MSTI 2026]. Austria’s research base is a real strength that capital, energy and labour costs make hard to monetise. Part II prices those costs, starting with what the state spends.

Part II: Why the Core Is Expensive

4 What the State Spends

The European core distributes more than half of output through the state, the east about 40% to 45%, and Ireland’s 21.0% reflects an inflated denominator rather than a lean state. Table 2 shows the 2025 ratios.

Table 2: Government expenditure, revenue and balance 2025 as a percentage of GDP

Country Expenditure Revenue Balance Debt
France 57.2% 52.1% −5.1% 115.6%
Austria 55.2% 51.0% −4.2% 81.5%
Belgium 54.2% 49.0% −5.2% 107.9%
Italy 51.2% 48.1% −3.1% 137.1%
Poland 50.9% 43.6% −7.3% 59.7%
Germany 50.5% 47.9% −2.7% 63.5%
Sweden 49.9% 48.6% −1.3% 35.1%
Luxembourg 49.1% 47.1% −2.0% 26.5%
Denmark 48.0% 50.9% +2.9% 27.9%
Hungary 47.3% 42.6% −4.7% 74.6%
Estonia 45.4% 43.4% −2.0% 24.1%
Netherlands 44.9% 43.3% −1.6% 44.4%
Romania 43.3% 35.4% −7.9% 59.3%
Czechia 43.2% 41.0% −2.1% 44.3%
Bulgaria 41.7% 38.1% −3.5% 29.9%
Lithuania 41.2% 39.4% −1.8% 39.5%
Ireland 21.0% 22.7% +1.8% 32.9%
EU-27 49.5% 46.4% −3.1% 81.7%
United States (general government) [PLACEHOLDER: IMF Fiscal Monitor 2026] [PLACEHOLDER] −5.8% (CBO, fiscal year 2026) 101% (federal, held by the public)

(Source: Eurostat, 2026c; CBO, 2026)

The CBO’s February 2026 baseline puts the US federal deficit at 5.8% of GDP in 2026 and 6.7% by 2036, debt held by the public rising from 101% to 120% and passing the 1946 record in 2030, and net interest growing from 3.3% to 4.6% of GDP, with 2.0 trillion dollars added after the tariff ruling (CBO, 2026). American dynamism is partly debt-financed; the European high-tax states fund their spending largely from current revenue.

The ratio buys different things. In Austria and Germany it covers health insurance for dependants without extra premium, a statutory pension, severance provision and fee-free university study; a US employer buys health cover privately at 9,325 dollars for a single person and 26,993 for a family in 2025, with employees paying 16% and 26% (KFF, 2025) [PLACEHOLDER: Eurostat gov_10a_exp, functional breakdown 2024].

Consolidation paths diverge within the core. Austria has been in an excessive-deficit procedure since 8 July 2025 with a 2028 deadline and expects a 4.2% deficit again in 2026 (Court of Audit, 2026); France keeps its surcharges while deficits exceed 5%; Romania raised VAT to 21% and dividend tax to 16% after a 7.9% deficit; Germany, at 2.7%, adopted a debt-financed investment programme and cuts corporate tax from 15% to 10% between 2028 and 2032 (Federal Ministry of Finance, 2025). Austria and France gain room only after 2028. Section 5 shows how that spending is collected and who bears it.

5 How It Is Paid For: The Tax Stack

An owner bears the integrated burden on distributed profit, the top rate on salary and the employer’s non-wage charges, not the headline corporate rate. On all three the European core matches the heaviest US states; on labour it exceeds every one of them. Table 4 summarises the stack (integrated burden = 1 − (1 − corporate tax) × (1 − dividend tax); Tax Foundation values where published, own calculations marked ≈).

Table 4: Tax and contribution stack 2025/26 by country and US state

Jurisdiction Corporate tax Dividend tax Integrated burden Top income-tax rate Employer charges VAT or sales tax
Austria 23% 27.5% 44.2% 55% (above 1 million euros, to end-2029) ≈29.5% 20%
Germany 30.1% incl. trade tax 26.375% 48.5% 47.5% incl. solidarity surcharge ≈21%, capped 19%
France 25% (+ surcharge from 1.5 billion euros turnover) 31.4% PFU 2026 + CEHR/CDHR 57.9% large firm; ≈54% SME owner 55.4% highest non-wage share in the EU (32.3%) 20%
Netherlands 25.8% 31% (box 2) ≈48.8% 49.5% moderate, capped 21%
Belgium 25% 30% 47.5% ≈53.5% high 21%
Sweden 20.6% 30% ≈44.4% ≈52% non-wage share 31.7% 25%
Denmark 22% 42% 54.8% 60.5% low (tax-financed) 25%
Italy 27.8% 26% 46.6% ≈47% high 22%
Ireland 12.5% (15% Pillar Two) 51% 57.1% 48% low 23%
Luxembourg 23.9% 21% 39.9% ≈45.8% low 17%
Czechia 21% 23% 39.2% 23% ≈34% 21%
Poland 19% (9% small firms) 19% ≈34.4% 32% + 4% solidarity levy ≈20% 23%
Lithuania 16% 15% 28.6% 32% low 21%
Hungary 9% + local business tax 15% 22.7% 15% ≈13% 27%
Estonia 0% retained / 22% distributed 0% 22% 22% 33% social tax 24%
Romania 16% (1% micro regime up to 100,000 euros turnover) 16% (2026) ≈29.4% 10% 2.25% (employee 35%) 21%
Bulgaria 10% 5% 14.5% 10% ≈19%, capped 20%
US federal (C-corp) 21% 23.8% (20% + 3.8% NIIT) 39.8% 37% FICA 7.65% + health ≈7,900 dollars —
Texas ≈0.75% franchise tax 0 ≈40% 37% + unemployment insurance 6.25% + local
Florida 5.5% 0 ≈43.1% 37% — 6% + local
Washington gross-receipts (B&O) tax 0 (capital-gains tax on exit) ≈40% + B&O 37% — 6.5% + local
California 8.84% 13.3% ≈54.7% 50.3% + SDI, unemployment insurance 7.25% (avg. 8.99%)
New York (state / NYC) 6.5–7.25% / + 8.85% city up to 10.9% / + 3.876% ≈48.8% / ≈56% 47.9% / 51.8% — 4% + local
Massachusetts 8% 5% (+ 4% above ≈1 million dollars) ≈48.3% 46% — 6.25%
Mississippi 5% ≈4% ≈45.8% ≈41% — 7%

(Source: Tax Foundation, 2025, 2026; OECD, 2026d; national tax authorities; Dutch, Belgian, Swedish, Italian and Polish personal rates and the Massachusetts and Mississippi brackets are unverified)

Seven rule changes are verified for 2025/26: Austria’s 55% band runs to end-2029 and reverts to 50% in 2030, with the family-fund contribution falling to 2.7% in 2028; Germany cuts corporate tax from 15% to 10% in 2028–2032; France raised the PFU to 31.4% on 1 January 2026 and extended the CDHR minimum tax and the large-company surcharge; Estonia stayed at 22% after the Riigikogu withdrew the 24% plan on 3 December 2025; Romania raised dividend tax to 16% and VAT to 21%; the US made the 37% top rate and 20% QBI deduction permanent in July 2025; and the OECD exempted US-parented groups from the Pillar Two income-inclusion and undertaxed-profits rules from 2026 (EY, 2026).

Table 5 shows what remains from 1 million euros of profit distributed to a resident owner at top rates (1 euro = 1 dollar; CDHR as 20% of reference income less income tax and CEHR; the pass-through case assumes a product business with sufficient payroll).

Table 5: Net result from 1 million euros of distributed profit

Jurisdiction Corporate tax Tax at owner level Net to owner Effective
Bulgaria 100,000 45,000 855,000 euros 14.5%
Estonia 220,000 on distribution 0 780,000 euros 22.0%
Romania (2026) 160,000 134,400 705,600 euros 29.4%
Texas, C-corp 210,000 188,020 601,980 dollars 39.8%
Florida, C-corp 253,450 177,680 568,870 dollars 43.1%
Austria 230,000 211,750 558,250 euros 44.2%
Germany (average trade tax) 300,600 184,467 514,933 euros 48.5%
France, SME owner 2026 250,000 ≈289,500 (PFU + CEHR + CDHR) ≈460,500 euros ≈54%
California, C-corp 279,800 267,190 453,010 dollars 54.7%
Texas, S-corp/LLC with 20% QBI 0 ≈252,000 ≈748,000 dollars ≈25%
California, S-corp 1.5% entity tax + 13.3% on full income — ≈600,000–620,000 dollars ≈38–40%

(Source: own calculation based on Table 4)

Austria sits between Florida and California, Germany level with Massachusetts and New York State, France level with California. The decisive US advantage for SMEs is pass-through taxation at about 25%, which no EU country offers; Austria’s Gewinnfreibetrag and Germany’s section 34a relief only defer tax at 28.25% falling to 25%.

Table 6 shows an 80,000-euro software developer (93,000 dollars), single, at 2026 rates with ±3 points of uncertainty (Austria: 20.98% employer and 18.07% employee contributions, ceiling 6,930 euros a month; Germany: ceilings 69,750 and 101,400 euros, additional health contribution 2.9%; US: 9,325-dollar single premium, 16% paid by the employee; KFF, 2025).

Table 6: Employer cost and net pay at 80,000 euros gross

Jurisdiction Total employer cost Employee net Net per euro of employer cost
Austria (14 salaries, Vienna) ≈103,700 euros ≈51,000 euros ≈49%
Germany (childless) ≈97,000 euros ≈48,000 euros ≈49%
France (cadre) ≈113,000 euros ≈52,000 euros ≈46%
Estonia ≈106,600 euros ≈62,000 euros ≈58%
Romania ≈81,800 euros ≈46,800 euros ≈57%
Bulgaria (capped contributions) ≈84,800 euros ≈69,000 euros ≈81%
Texas (incl. ≈7,900 dollars employer premium) ≈108,300 dollars ≈72,800 dollars ≈67%
California ≈109,000 dollars ≈67,500 dollars ≈62%

(Source: own calculation based on LexisNexis, 2026; Osborne Clarke, 2026; KFF, 2025)

A no-income-tax US state delivers about 1.4 times the net pay per euro of labour cost of Austria or Germany, even after private health insurance. OECD Taxing Wages 2026 confirms it: the tax wedge on a single average earner is 52.5% in Belgium, 49.3% in Germany, 47.2% in France and 47.1% in Austria against 35.1% for the OECD and about 30% in the US, and Austria’s marginal wedge is 58.3% (OECD, 2026d).

Verdict: owner-level capital tax is not the core’s distinctive problem, since Austria’s 44% sits near the US population-weighted average of 47% and below California; labour taxes are, at 15 to 20 points above the US with no state close; the top rates of 55% in Austria and France and 60.5% in Denmark punish founders who take salary; and the missing pass-through regime is the largest structural gap for SMEs. In Central and Eastern Europe the constraint is income, capital depth and, in Romania, fiscal instability, not tax. The tax stack is only part of the operating cost; energy and labour add the rest.

6 What Else a Firm Pays: Energy and Labour

Industrial electricity in Germany, Austria, Denmark and Italy costs three to four times the Texan price; in Sweden and Finland it costs less than in California. Eurostat’s second-half 2025 non-household price (band IC, excluding recoverable taxes) was 18.37 cents per kilowatt-hour for the EU, 22.64 in Germany, 25.52 in Ireland and 7.48 in Finland (Eurostat, 2026d); the all-in 2025 ranking puts Ireland, Germany, Denmark, Hungary and Italy at 26 to 29 cents, Austria at 24.1, Romania 21.6, France 18.9, Bulgaria 16.1, Sweden 12.1 and Finland 9.6 (Apiar, 2026). EIA industrial prices for January–June 2026 are 8.89 US cents nationally, 6.65 in Texas, 7.09 in Washington and 19.88 in California (EIA, 2026); the IEA puts EU energy-intensive industry above twice the US level and ACER puts EU gas at three times the US price (IEA, 2026; ACER, 2026).

The 2026 Hormuz shock widened the gap where it was widest. After the strikes on Iran on 28 February 2026 the TTF rose almost 50% to 47.94 euros per megawatt-hour on 3 March and reached 80.32 euros on 11 September, Henry Hub rose about 13% in the first week against 70% for TTF, and euro-area energy inflation hit 14.3% in August (CEPR, 2026; CNBC, 2026). Germany, Italy, Austria and Hungary are most exposed; Sweden, Finland and France least.

Table 7: Labour cost per hour 2025 in the business economy

Group Country Euros per hour
Core Luxembourg 56.8
Denmark 51.7
Netherlands 47.9
Austria 46.3
Germany 45.0
France 44.3
Ireland 44.2
Sweden 43.1
Middle EU average 34.9
Italy 32.0
East Estonia 21.1
Czechia 19.8
Poland 19.1
Lithuania 17.8
Hungary 15.2
Romania 13.6
Bulgaria 12.0

(Source: Eurostat, 2026e)

Austria’s manufacturing labour cost of 51.3 euros is second only to Denmark’s. The east is closing the gap by five to eight points a year, with wages up 13.1% in Bulgaria and 10.6% in Romania in 2025, but a development seat in Vienna or Munich still costs 2.3 to 3.5 times one in Bucharest, Kraków or Sofia, which is why Austrian and German firms run centres there [PLACEHOLDER: BLS employer costs 2026]. Costs can be priced; Section 7 turns to the rules that shape them.

7 How the Rules Are Made and Enforced

European governments optimise for stability and precaution, American governments for speed, and three indices measure the result: the Corruption Perceptions Index (CPI) for perceived public-sector corruption, the OECD Product Market Regulation indicator (PMR) for the competition-friendliness of rules, and the IMD ranking for the ability to attract investment. Table 3 sets them side by side.

Table 3: Corruption perception 2025, IMD competitiveness rank 2026 and product market regulation 2023/24

Country CPI 2025 (0–100) CPI 2024 IMD rank 2026 (of 70) PMR 2023/24
Denmark 89 90 [PLACEHOLDER] Nordic average about 1.0
Finland 88 88 [PLACEHOLDER] Nordic average about 1.0
Sweden 80 80 [PLACEHOLDER] among the three simplest regimes
Switzerland 80 81 3 (government efficiency: 1) [PLACEHOLDER]
Luxembourg 78 81 [PLACEHOLDER] [PLACEHOLDER]
Netherlands 78 78 8 [PLACEHOLDER]
Germany 77 75 23 (2025: 19) [PLACEHOLDER]
Estonia 76 76 [PLACEHOLDER] rank 6
Ireland 76 77 7 among the three simplest regimes
Austria 69 67 29 (2025: 26; government efficiency 2025: 40) [PLACEHOLDER]
Belgium 69 69 32 [PLACEHOLDER]
France 66 67 36 [PLACEHOLDER: 2018 value 1.57]
Lithuania 65 63 [PLACEHOLDER] rank 1
United States 64 65 10 worse than OECD average
Czechia 59 56 [PLACEHOLDER] [PLACEHOLDER]
Italy 53 54 45 [PLACEHOLDER]
Poland 52 53 [PLACEHOLDER] [PLACEHOLDER]
Romania 45 46 [PLACEHOLDER] not an OECD member
Bulgaria 40 43 [PLACEHOLDER] not an OECD member
Hungary 40 41 [PLACEHOLDER] [PLACEHOLDER]

(Source: Transparency International, 2026; IMD, 2026; OECD, 2026c; Lithuanian Ministry of Economy and Innovation, 2026)

On corruption the core and the north are cleaner than the US: Germany scores 77 (10th), Austria 69 (21st), the US 64 (29th) and twelve points lower than in 2015 (Transparency International, 2026). Bulgaria, Hungary and Romania sit at 40 to 45, a range Transparency International classes as a serious problem, so the periphery’s low tax rates carry a legal-certainty discount.

On competitiveness the order reverses. IMD 2026 ranks Switzerland third with first place in government efficiency, Ireland seventh, the Netherlands eighth and the US tenth, while Germany fell from 19th to 23rd, Austria from 26th to 29th, and France sits 36th (IMD, 2026); Austria’s government efficiency ranked 40th already in 2025. On formal regulation Lithuania leads the 2023/24 PMR, Sweden and Ireland are among the three simplest regimes, Estonia rose to sixth, the Nordics average 1.0 on a 0–6 scale, and the US scores worse than the OECD average on administrative burdens, licensing and start-up requirements (Lithuanian Ministry of Economy and Innovation, 2026; OECD, 2026c). B-READY 2025 covers neither the core economies nor the US but scores Romania 86 on business entry and 52 on insolvency (World Bank, 2025).

The indices capture rules, not execution, which is where Austria’s burden lies: compulsory Chamber of Commerce membership with a levy, licensing under the Gewerbeordnung, employer charges of about 29.5% on every gross wage (social insurance 20.98%, severance fund 1.53%, family fund 3.7% plus surcharge, municipal payroll tax 3%; LexisNexis, 2026), and a record 6,857 insolvencies in 2025, 19 a day, concentrated in trade, construction and hospitality (KSV1870, 2026). Germany adds compulsory chamber membership, trade tax and slow permitting (DIHK, 2025).

Brussels has conceded the point three times. Only 15.7% of the Draghi report’s 383 recommendations were fully implemented by July 2026, 41.3% counting partial progress (EPIC, 2026). The Digital Omnibus (Regulation (EU) 2026/1744) postponed the AI Act’s high-risk obligations to December 2027 and August 2028 (European Commission, 2026a). EU Inc. (COM/2026/321), an optional EU-wide company form with a pan-EU employee share scheme, received the Council’s position on 23 October 2026 with the share scheme kept and tax harmonisation dropped (Council of the European Union, 2026).

The American alternative is regulation by court ruling with a short lifespan. The Supreme Court struck down the IEEPA tariffs on 20 February 2026 after 175 billion dollars had been collected, cutting the trade-weighted average tariff from 15.3% to 8.3% (Global Trade Alert, 2026); a Section 122 surcharge replaced them within days, expired on 24 July, and gave way the same day to an uncapped Section 301 action covering more than 80 economies (Yale Budget Lab, 2026). The EU’s 15% ceiling has applied since 1 July 2026, with steel and aluminium at 50% [PLACEHOLDER: Official Journal]. European investors cannot lose a trade regime overnight; American investors can. Part III follows where capital and people go under these conditions.

Part III: Where Capital and Talent Go

8 Capital Availability

Europe lacks risk capital, not founders, because its savings sit in banks, insurance and pay-as-you-go entitlements rather than equity. Pension assets exceeded GDP at end-2024 in Denmark (206.4%), Switzerland (166.9%), the US (153.3%), the Netherlands (150.9%) and Sweden (115.8%), while Germany holds about 7% of GDP (OECD, 2025a; Apollo, 2026a); US pension and household investment assets reach almost 475% of GDP and market capitalisation over 200%, with Germany, France, Italy and Spain far below (Apollo, 2026b) [PLACEHOLDER: Austria pension assets, % of GDP].

The consequence is a self-sustaining venture gap. European pension funds allocate 0.018% of assets to venture capital against 1.9% in the US (Jacques Delors Centre, 2024), contributed only 0.6 billion euros or 5% of new European venture fund capital in 2023 (CEPS-ECMI, 2025), and European funds return 8.6% a year against 14.6% (CEPR, 2025), so European venture volume stays at about 22% of the US level.

Two rules reinforce it. Pillar Two now exempts US-parented groups while European groups face the full regime (EY, 2026), and national employee-equity schemes (Austria’s 2024 start-up scheme, Germany’s section 19a, France’s BSPCE, Sweden’s QESO) stop at the border and trail US incentive stock options and QSBS [PLACEHOLDER: 2026 parameters]. That is why the pan-EU share scheme in EU Inc. matters more than any rate cut.

The core’s Mittelstand is built for this structure: bank-financed family firms that fail less often and scale less often. Austria’s 2025 insolvency wave hit trade, construction and hospitality, not technology (KSV1870, 2026); European insolvency regimes are slower than Chapter 11; and the New York listings of Arm, Spotify, Klarna and UiPath show where pension money, option regimes and cost of capital align [PLACEHOLDER: Eurostat SBS; ECB SAFE 2026; EIB Investment Survey 2025; Savings and Investments Union status]. Thin capital shows in investor sentiment, which Section 9 measures.

9 Investment Climate in Autumn 2026

Sentiment is subdued on both sides in autumn 2026, for different reasons. The euro-area Economic Sentiment Indicator fell to 97.9 in September from 98.4, its first decline since April, on oil prices and the stalled talks to end the US–Iran war; July country values were 97.2 for Germany, 95.9 for France, 99.6 for Italy and 104.3 for Spain (European Commission, 2026b) [PLACEHOLDER: ESI, ifo and WIFO October 2026]. In the US the ISM manufacturing index stood at 54.5 in September with new orders at 55.3 (ISM, 2026), while NFIB small-business optimism fell to 97.9, just below its 52-year average, with uncertainty at 90 against a mean of 68 and labour availability (26%) and inflation (20%) the top problems (NFIB, 2026). Industry invests; small firms wait.

Venture capital is where the temperature differs. US investors deployed a record 515.8 billion dollars in the first three quarters of 2026, 82.7% in AI, with OpenAI and Anthropic raising over 200 billion in the first half (PitchBook-NVCA, 2026); Europe raised 44 billion euros in the first half, up 27% on an annual basis, 60.3% in AI (PitchBook, 2026), and Dealroom expects a record 10.5 billion dollars for European defence start-ups (Defense News, 2026). The IMF flags the concentration as a risk: if AI productivity disappoints, US GDP ends 1.8% below baseline by 2028 (IMF, 2026).

Europe’s investment need of 1.2 trillion euros a year (Draghi, 2025) has no matching volume; German fiscal expansion and defence demand are the only sources that can narrow it within five years [PLACEHOLDER: GFCF 2025 by country; UNCTAD WIR 2026; EIB survey 2025]. Where the capital that does flow ends up is the subject of Section 10.

10 The Technology Market by Country

US firms took about 75% (194 billion dollars) of global AI venture capital in 2025, the EU-27 about 6% (15.8 billion) (OECD, 2026a). European venture totals for 2025 differ by method, not magnitude: 66.2 billion euros (PitchBook), 72 billion (Tech.eu), 85.3 billion dollars (KPMG), against about 339 billion dollars in the US with 65.6% in AI. The UK led Europe with 21.5 billion euros, then Germany 11.5, France 8.7 and Sweden about 3.2 billion dollars (Tech.eu, 2026; Dealroom, 2026); the largest 2026 rounds were Nscale (2.9 billion euros), Isomorphic Labs (2.1 billion dollars) and Neura Robotics (1.4 billion euros).

Capital concentrates in a few clusters. Paris: Mistral raised 1.7 billion euros at 11.7 billion in September 2025 and 3 billion at over 21 billion in September 2026, Europe’s largest equity round. Munich: Celonis (1.8 billion dollars total), Helsing (1.8 billion Series E at 18 billion, July 2026), and Germany’s 3.5 billion dollars of defence venture capital, now 15% of the EU total. Stockholm: Lovable, Legora, Klarna and Spotify, the EU’s highest unicorn density. Amsterdam: ASML and Adyen. The second circle is Austria (Bitpanda, GoStudent, TTTech; an order of magnitude below Munich, no foundation-model cluster, the highest labour costs in its peer group [PLACEHOLDER: AT venture volume 2025/H1 2026]), Estonia (Bolt, Wise; zero tax on retained profit), Romania (UiPath now US-based; Bitdefender; IT tax exemption removed in 2025) and Poland (the largest engineering pool and nearshoring hub).

Infrastructure follows the same pattern. US hyperscalers hold about 70% of Europe’s cloud market, European providers 15%, down from 29% in 2017 (Synergy Research, 2026). Intel cancelled its 30-billion-euro Magdeburg fab on 24 July 2025; ESMC Dresden (TSMC 70%; 28–12 nm; 40,000 wafers a month) targets production in late 2027; a Chips Act 2.0 (COM(2026) 504) awaits adoption in 2027. Europe attracts mature nodes, not frontier logic. Defence is the exception: European military spending rose 14% to 864 billion dollars in 2025, Germany 24% to 114 billion, while US spending fell 7.5% (SIPRI, 2026). The AI Act’s high-risk rules are deferred to 2027/28; US federal AI policy is deregulatory. Companies form in Europe and scale or list in the US for the reasons Sections 5, 6 and 8 quantified. People move along the same routes, with one reversal that Section 11 describes.

11 Migration and Talent

Europe now takes in more people than the US. EU net migration was 2.05 million in 2025, offsetting a natural decline of 1.35 million (Eurostat, 2026f); US net immigration was 410,000 in calendar 2025 on CBO figures, 1.3 million for July 2024–June 2025 on Census figures after 2.7 million the year before, and about 321,000 projected to mid-2026, possibly turning negative (CBO, 2026; Census Bureau, 2026).

The US has made itself costlier for skilled entrants. A 100,000-dollar fee on certain H-1B petitions has applied since 21 September 2025, was vacated and blocked by federal courts in June and September 2026, yet was extended by proclamation to September 2027 pending appeal, and the lottery has been wage-weighted since 27 February 2026 (TheJobSeek, 2026) [PLACEHOLDER: USCIS H-1B approvals FY2026]. Europe’s counter-offer, Choose Europe for Science with nearly 900 million euros by January 2026 and more than 100 national schemes (European Commission, 2026c), plus Germany’s Chancenkarte and the reformed Blue Card, is blunted by language, salaries and slow administration; Austria’s Red-White-Red Card competes with a 47% tax wedge against about 23% in Switzerland and 30% in the US [PLACEHOLDER: card volumes 2025/26].

Inside Europe, skilled workers move west while firms move work east to Bucharest, Cluj, Kraków, Brno and Sofia; inside the US, people and capital move from California and New York to Texas and Florida while frontier AI stays in California. Demographics settle the European question: the old-age dependency ratio of 34.5% in 2025 rises to about 60% by 2100 (Eurostat, 2026g), so immigration is a necessity, and the open question is whether Europe attracts engineers or only carers. Part IV combines these forces into scenarios.

Part IV: Outlook

12 Forecast to 2031

The likely outcome is managed divergence: the US keeps frontier AI and the hyperscaler layer, Europe holds 20% to 30% of US venture volume and builds strength in defence, industrial AI and sovereignty niches, and the growth gap of about one point a year persists. Four factors can shift it: a US AI-investment correction, US immigration restrictions, European energy prices, and whether Brussels delivers EU Inc. and capital-market integration.

Institutional projections: IMF (October 2026) 2.3% US growth in 2026 and 2027, 1.1% and 1.2% for the euro area, 1.3% for Germany and 0.4% for France in 2026; CBO about 1.8% US growth after 2026 with deficits averaging 6.1% of GDP; WIFO 0.9% for Austria in 2026. Table 8 gives the scenarios; probabilities are the author’s judgment and the downside cases overlap.

Table 8: Scenarios for the technology market to 2031

Scenario Probability Europe United States
Base: managed divergence about 55% Growth 1% to 1.5%. Venture capital at 20% to 30% of US volume. Strong in defence, industrial AI, robotics, sovereign cloud. AI Act high-risk rules land softened in 2027/28. EU Inc. usable from about 2028. Growth about 2%. More than 70% of global AI capital. Tariffs settle at 10% to 15%. Debt above 108% of GDP by 2030. Frontier AI in California, firms moving to Texas and Florida.
Europe upside: Draghi delivered about 20% EU Inc. with its share scheme, progress on the Savings and Investments Union, defence and energy spending crowd in private capital, German fiscal expansion works. Venture gap narrows to 35% to 40%. Hormuz resolved, TTF at 30 to 35 euros, industrial power only twice the Texan price. Return migration rises. As base or weaker.
US downside: AI investment correction about 20% Relative winner, though dollar-funded European AI firms are hit too. AI investment and markets correct (IMF: US GDP 1.8% below baseline in 2028). Falling immigration tightens labour. Higher interest on debt near 120% of GDP; dollar safe-haven status questioned.
Europe downside: energy-locked stagnation about 15% to 25% Hormuz disruption into 2027 with TTF above 70 euros, ECB keeps raising, energy-intensive industry shrinks, Austrian insolvencies rise, France raises taxes, Romania misses its deficit path, founders relocate to the US and Switzerland. Relative winner.

(Source: own presentation)

In the base case Germany recovers to 1% to 1.5% as fiscal expansion and the 2028–32 tax cut take hold; Paris stays the EU’s AI centre under surcharges that last to about 2028; Austria grows about 1% and cannot cut tax before 2029, remaining a liveable but expensive secondary hub that exports its scale-ups to Munich and the US; Poland, Czechia and the Baltics grow 2.5% to 3.5% on nearshored engineering; Romania consolidates below 2% to about 2027; and the US keeps California for frontier AI while Texas and Florida gain firms and people.

Three conclusions follow. Bet on US scale and European niches: defence, industrial and robotics AI, semiconductor equipment and software for regulated sectors. Watch European capital integration, not tax rates, since the EU Inc. trilogue and the capital-markets agenda say more about 2031 than any growth forecast and core tax rates will not fall before 2028–2030. Treat US talent restrictions as the largest self-inflicted risk: sustained H-1B tightening alongside scaling European programmes could make return migration Europe’s first structural gain in two decades, a hypothesis to test rather than a forecast.

For an Austrian founder in 2026: keep the operating company in Austria for subsidies and research talent, take profit as dividends (44%) not salary (up to 55%), build engineering in Poland, Romania or the Baltics, raise growth rounds through Munich or US vehicles, and review the structure in 2030.

13 Conclusion

The gap between Europe and the United States is a gap in dynamics that the European core pays for in three currencies: labour taxes 15 to 20 points above the US, industrial electricity at three to four times the Texan price, and savings held in banks and pay-as-you-go pensions rather than equity. Owner-level capital taxes, the usual suspect, match the high-tax US states and explain little. The low-tax periphery does not escape the pattern, since its advantage is cost rather than capital and its institutions carry a risk discount.

Three things decide 2031. The EU Inc. trilogue and the capital-markets agenda determine whether European savings reach European firms. The Hormuz energy outcome determines whether the gas-price-set markets of Germany, Austria, Italy and Hungary keep their industry. The US immigration stance determines whether the one flow that has turned in Europe’s favour, talent, becomes structural. Tax rates in the core will not move before 2028 to 2030, so none of this depends on them.

14 Sources and Items to Verify

Check these against primary tables before publication; they rest on secondary sources, conflict between sources, predate 2024 or are own calculations.

  • IMF growth figures of October 2026 (US 2.3%, euro area 1.1% to 1.2%) come from news coverage; reconcile with the WEO tables.
  • Euro area 2025: 1.4% (IMF update January 2026) against 1.2% (coverage October 2026).
  • Policy rates (Fed 3.75% to 4.00%, ECB deposit rate 2.50%) and the 2026 rate rises rest on Schwab, CNBC and market commentary.
  • Austria GDP per head in purchasing-power parities: World Bank 63,970 dollars against IMF about 76,000 dollars; growth 2025: 0.8% (Eurostat) against 1.0% (IMF).
  • US general-government expenditure ratio 2025 on the IMF or OECD definition is missing (Table 2).
  • Functional breakdown of government expenditure by country (Eurostat gov_10a_exp) is missing.
  • IMD ranks 2026 for Denmark, Sweden, Finland, Luxembourg, Estonia, Lithuania, Czechia, Poland, Hungary, Romania and Bulgaria and OECD PMR country values 2023/24 for Austria, Germany, France and the remaining countries are missing (Table 3).
  • EU–US tariff ceiling of 15% since 1 July 2026 and the Section 122 rate (10% or 15%) rest on trade-advisory sites; check against the Official Journal.
  • US venture capital 2025: 339 billion dollars (PitchBook-NVCA) against 320 billion dollars (NVCA yearbook); Europe 2025: 66.2 billion euros (PitchBook), 72 billion euros (Tech.eu), 85.3 billion dollars (KPMG) on different definitions.
  • Research intensity (EU 2.24% for 2024, US about 3.5%) rests on 2024 data and older OECD series.
  • Productivity per hour worked 2024 by country (OECD Productivity Database) is missing for Austria, Germany, France, the Netherlands, Sweden, Poland and Romania.
  • US industrial electricity prices 2026 refer to January to June; the Apiar series partly includes recoverable taxes.
  • Integrated burdens and net salaries marked ≈ are own calculations at top rates, without church tax, with average trade tax and 1 euro = 1 dollar; Dutch, Belgian, Swedish, Italian, Polish and Swiss personal rates and the Massachusetts and Mississippi brackets are unverified.
  • Austria’s pension assets as a percentage of GDP, ECB SAFE survey 2026, EIB Investment Survey 2025, Eurostat SME shares, BLS employer costs, current parameters of the employee-equity regimes, status of the Savings and Investments Union, ESI and ifo October 2026, gross fixed capital formation 2025, UNCTAD foreign direct investment 2025, H-1B approvals 2026, Red-White-Red Cards and opportunity cards were not verified (see placeholders in the text).
  • Mistral’s total funding is given by Dealroom as 6.2 billion dollars, which looks low against the two reported rounds.
  • EU old-age dependency ratio 2100: 59.7% or 61.9% depending on the Eurostat projection vintage.
  • Scenario probabilities in Table 8 are the author’s judgment.

References