Every 2026 number in this Europe forecast was wrong

A research report has been circulating since January that applies the Strauss–Howe generational theory to Europe. Its argument is clean and its conclusion is dramatic: Europe entered a Fourth Turning in 2008, the crisis climaxes around 2030, and by the mid-2030s the continent will have been remade — federalised, fragmented, or split into a core and a periphery.

I checked its numbers. Every 2026 indicator in it was wrong, and four of them were wrong in the opposite direction to the argument they were supporting.

That is not a reason to dismiss cyclical history. It is a reason to be careful about what a forecast is actually resting on, because the failure mode here is instructive and extremely common: a framework that is difficult to falsify, paired with data that nobody re-checked, produces a conclusion that feels evidenced without being evidenced.

What the framework says

William Strauss and Neil Howe proposed that Anglo-American societies move through recurring cycles of roughly 80 to 100 years, each containing four “turnings” of about 20 to 25 years: a High of strong institutions, an Awakening of values revolt, an Unraveling of institutional decay, and a Crisis in which the old order is destroyed and rebuilt. Four generational archetypes rotate through these turnings in fixed order — Prophet, Nomad, Hero, Artist — each born in one turning and coming of age in the next.

The report I read got the sequence wrong. It has Prophets “born during Crisis/High.” Prophets are born during a High. That sounds pedantic until you notice that the fixed archetype sequence is the theory — it is the mechanism by which turnings are supposed to generate the next turning. Break the sequence and you have a mood board, not a model.

The dating was also compressed. Howe's own 2023 restatement puts the resolution of the current Fourth Turning in the early-to-mid 2030s. The report moved it to 2030 without saying why, which conveniently placed the climax inside its own forecast window.

The numbers

Here is where it stops being a question of interpretation. The report published a table of “2026 Q1” indicators with crisis thresholds attached, presented as a live dashboard for validating the thesis.

Trust in the EU was given as 45% and falling. Standard Eurobarometer 105, fielded in spring 2026, puts it at 51% — up three points on autumn 2025 and near an eighteen-year high. Support for the euro is at 74% across the Union and 82% in the euro area, the highest since the currency was introduced. Eighty-one per cent back a common defence and security policy.

The Italy–Germany ten-year spread was given as 150 basis points, the classic thermometer for eurozone stress. It hit 59 basis points in January 2026 and sat around 75 in May. For scale, it went above 500 in the 2011–12 crisis.

EU unemployment was not tracked at all. It was 6.0% in June 2026, at or near a record low. The report's own confidence interval put the tenth percentile floor for peak unemployment at 11% — that is, it treated a near-doubling of European joblessness as the optimistic case.

A framework built to detect transformation will systematically over-predict it. That is not a flaw in Strauss and Howe. It is a flaw in using them as an instrument rather than a hypothesis.

None of this means Europe is fine. Fertility is at roughly 1.36 and falling. Radical-right parties are in or propping up governments across the continent. Rearmament is proceeding at a scale without post-1950s precedent, under open doubt about American guarantees. The Iran conflict pushed euro area inflation to 3.2% in May, the euro area contracted 0.2% in the first quarter, and in June the ECB raised rates for the first time in nearly three years.

But that is a different stress profile from the one the report forecast. It predicted a deflationary debt crisis and institutional delegitimation. What has arrived is a stagflationary energy shock layered onto a defence build-out, at a moment when European institutions are more trusted than they have been since 2007. Getting the direction of your central variable wrong is not a rounding error.

The structural problem

Underneath the data errors sits something harder to fix.

The report's probability matrix ran three phases: 2026–2029, 2028–2031, 2030–2035. Each overlaps the next by two years, so a 2029 event falls in two phases at once and the phase probabilities cannot be interpreted. The third phase had no probabilities in it at all — the column was empty — yet the document reported an aggregate figure of 55% for that phase. It described its method as “Monte Carlo-style reasoning with historical base rates.” No simulation was run and no base rates were cited.

And the forecasts themselves were mostly not forecasts. “ECB credibility crisis — 45%.” What resolves that? “AI Singularity/AGI emergence — 8%.” Measured how, by whom, by when? A probability attached to an undefined event is a mood expressed in decimals.

One entry deserves singling out. Among the “variables that would increase crisis severity,” the report listed a Trump or isolationist US presidency in 2025–2029, adding fifteen percentage points to geopolitical crisis risk. That administration took office in January 2025. Treating a realised state of the world as an upside risk double-counts it across the entire matrix.

What survives

Three things, and they are worth keeping.

Generational forgetting is real. Roughly eighty years is the interval at which direct institutional memory of a catastrophe is extinguished — the last people who ran the last crisis stop being in the room. This is the one part of the theory with a plausible causal mechanism, and it does not require the four-phase apparatus to be true.

Institutional lifespan is real. The arrangements that govern Europe were built between 1945 and 1957 and last substantially reformed at Maastricht. Systems designed in one crisis calcify and eventually fail to bend. Reasoning that assumes their permanence is fragile reasoning, whatever you think of saecula.

Enlargement is a forcing function. A Union of thirty-plus members cannot operate under current unanimity rules, and everyone involved knows it. Something gives in the next decade. That is not a cyclical prediction; it is arithmetic.

The correction that matters most

The original gave “muddling through” — crisis managed, incremental reform, no transformation — a 10% chance by 2035. I put it at 35%, and it is the single largest revision in the review.

Institutions of this age and density very rarely dissolve and very rarely refound themselves. They accrete workarounds. The realistic bad outcome for Europe is not a collapsed Union but an intact one that cannot decide: a formally complete institution, systematically obstructed, from which a coalition of the willing quietly exits into parallel arrangements. That is far duller than a constitutional convention, and far more likely.

Winter metaphors are seductive because they promise that the frost is going somewhere. Sometimes it just sits there.

The full review runs to a corrections log of 33 items, a rebuilt probability appendix in which every proposition carries a resolution criterion and a date, and a verified indicator dashboard. Data current to mid-2026 and sourced to Eurobarometer, Eurostat, the ECB and the Italian Ministry of Economy and Finance.