Data story · 30 August 2026
Europe Is Flying Again.
But the Old Hub Map Is Not.
Commercial aviation nearly recovered its flight count. Its connecting network did not.
The convenient explanation
Flights are almost back. Therefore, the old hub system must be almost back too.
It is a plausible inference. It is also wrong.
The EU recorded 6,905,681 commercial flight movements in 2025, 3.8% more than in 2024 and 2.0% fewer than in 2019. One movement is one take-off or landing. It is not a passenger, a seat, or an emission.
Flights recovered to 98. Hub connectivity recovered to 88.
Every measure is indexed to its own 2019 level. The common scale makes the recovery gap visible; it does not make the underlying concepts identical.
Exact values, definitions, and download
| Measure | Index | Change | Definition |
|---|---|---|---|
| Commercial flight movements | 97.97 | −2.0% | All EU commercial IFR flight movements; one take-off or landing is one movement. |
| Direct connectivity | 95.00 | −5.0% | Direct destinations weighted by flight frequency. |
| Indirect connectivity | 89.00 | −11.0% | Destinations reachable via a hub, weighted by connection time and detour. |
| Air connectivity | 91.00 | −9.0% | Direct plus indirect connectivity. |
| Hub connectivity | 88.00 | −12.0% | Connecting flights facilitated at a hub, weighted by time and detour quality. |
Sources: Eurostat/Eurocontrol and SEO Amsterdam Economics / ACI Europe. Calendar year 2025.
Act I · Read the unit
Movement recovery and network recovery are different claims.
Across the 476 airports in the ACI Europe and SEO study, direct connectivity remained 5% below 2019. Indirect connectivity was 11% lower. Hub connectivity—the quantity and quality of connecting opportunities an airport can facilitate—was still 12% below.
Act II · The map changes
The airport evidence rejects a simple hub-decline story.
Istanbul reached +13% direct and +59% hub connectivity. Athens reached +24% and +92%. Frankfurt remained 11% below in direct and 21% below in hub connectivity.
Rome is the useful counterexample: direct connectivity rose 3%, while hub connectivity fell 26%.
Hub power did not disappear. It moved.Hub power did not disappear. It moved.
The twenty leading direct-connectivity airports occupy four different recovery states. Position compares change, not score size.
Exact top-twenty values and download
| Rank | Airport | Direct change | Hub change | Gap |
|---|---|---|---|---|
| 1 | IST | +13% | +59% | +46 pp |
| 2 | AMS | −4% | −17% | −13 pp |
| 3 | LHR | −2% | −5% | −3 pp |
| 4 | FRA | −11% | −21% | −10 pp |
| 5 | CDG | −7% | −10% | −3 pp |
| 6 | MAD | −3% | −2% | +1 pp |
| 7 | BCN | +1% | +5% | +4 pp |
| 8 | MUC | −15% | −12% | +3 pp |
| 9 | FCO | +3% | −26% | −29 pp |
| 10 | ATH | +24% | +92% | +68 pp |
| 11 | PMI | +11% | +93% | +82 pp |
| 12 | LGW | −7% | −23% | −16 pp |
| 13 | SAW | +14% | +75% | +61 pp |
| 14 | CPH | −1% | +7% | +8 pp |
| 15 | VIE | −8% | −32% | −24 pp |
| 16 | DUB | +8% | +30% | +22 pp |
| 17 | ZRH | −2% | −5% | −3 pp |
| 18 | ORY | −4% | −44% | −40 pp |
| 19 | LIS | +4% | +8% | +4 pp |
| 20 | AYT | +29% | +139% | +110 pp |
Source: SEO Amsterdam Economics / Airports Council International - Europe (ACI EUROPE), 2025 annex.
One ranking. Four different signals.
Scores are weighted network measures. Compare rank and change—not the direct score with the hub score.
Exact scores, changes, and download
| Rank | Airport | Direct score | Direct change | Hub score | Hub change |
|---|---|---|---|---|---|
| 1 | IST | 5059 | +13% | 66021 | +59% |
| 2 | AMS | 4691 | −4% | 48119 | −17% |
| 3 | LHR | 4588 | −2% | 31968 | −5% |
| 4 | FRA | 4523 | −11% | 61977 | −21% |
| 5 | CDG | 4436 | −7% | 42547 | −10% |
| 6 | MAD | 3833 | −3% | 22501 | −2% |
| 7 | BCN | 3466 | +1% | 4473 | +5% |
| 8 | MUC | 3428 | −15% | 30068 | −12% |
| 9 | FCO | 3381 | +3% | 10957 | −26% |
| 10 | ATH | 2936 | +24% | 6754 | +92% |
Source: SEO Amsterdam Economics / Airports Council International - Europe (ACI EUROPE), 2025 annex.
Signature interaction · Put the hub story on trial
Choose an airport. Judge the same evidence.
Flights, direct links, hub connectivity, seats, and passengers remain separate. The card changes; the evidence boundary does not.
Frankfurt/Main FRA
Direct and hub connectivity are both below 2019. The old network is not back.
Medium airports regained direct links. Small airports did not.
Direct access and indirect network recovery do not move together. Focus, hover, or tap a segment for its exact comparison.
Exact segment values and download
| Segment | Direct | Indirect | Total air |
|---|---|---|---|
| Major | −3% | −6% | −5% |
| Mega | −3% | −6% | −5% |
| Large | −4% | −12% | −10% |
| Medium | +2% | −9% | −5% |
| Small | −8% | −16% | −13% |
Source: ACI Europe Airport Industry Connectivity Report 2025, Chart 5. Calendar year 2025 versus 2019.
Act IV · Selective bypass
Medium-airport gains do not amount to a general bypass revolution.
Medium airports were 2% above 2019 in direct connectivity, consistent with selective new access. Their indirect connectivity remained 9% lower. Small airports remained below 2019 on every measure.
Act V · Airline strategy
The clearer structural split is between business models.
Direct connectivity supplied by low-cost carriers was 19% above 2019. Full-service-carrier connectivity was 15% below. Aircraft technology may matter; airline network strategy moved first in the public evidence.
The clearer split is between airline business models.
All-direct connectivity changed in opposite directions for low-cost and full-service carriers. Focus, hover, or tap a row for its exact change.
Exact business-model values and download
| Model | All-direct change | Intra-Europe share 2019 | 2025 |
|---|---|---|---|
| Low-cost carriers | +19% | 32% | 39% |
| Full-service carriers | −15% | — | — |
| All carriers | −5% | — | — |
Source: ACI Europe 2025 report, Chart 20 and page 27. Calendar year 2025 versus 2019.
Act VI · Frankfurt on the evidence bench
A flight is not a seat. A seat is not a passenger.
The objection to counting flights is correct. Fewer flights can carry more seats, and fuller aircraft can carry more passengers. At Frankfurt, the capacity repair does not reverse the result.
Passenger flights remained 11.7% below 2019. Seats per flight rose only 0.7%. Passenger numbers remained 10.6% lower.
Frankfurt’s gap is still mainly a frequency gap.
A flight is not a seat. A seat is not a passenger. Four measures keep the capacity repair honest.
Exact capacity values, definitions, and download
| Airport | Passenger flights | Seats / flight | Occupancy | Passengers |
|---|---|---|---|---|
| Frankfurt/Main | −11.7% | +0.7% | +0.5 pp | −10.6% |
| Madrid-Barajas | +2.3% | +8.1% | +2.5 pp | +13.8% |
| Munich | −19.5% | +7.2% | +3.7 pp | −9.6% |
| Rome Fiumicino | +1.4% | +7.6% | +5.6 pp | +17.0% |
| Copenhagen | −2.8% | +5.5% | +3.4 pp | +7.2% |
Source: Eurostat avia_tf_aca; author calculations. Athens is excluded because 2019 available-seat data are not reported.
The aircraft claim is plausible. It is not identified.
The public evidence describes the shift. It cannot isolate the A220 or A321XLR as its cause.
Exact evidence ledger and download
| Test | Public source | Latest year | Aircraft causality? | Status |
|---|---|---|---|---|
| Airport direct and hub connectivity in 2025 | ACI/SEO NetScan annex | 2025 | No | Observed |
| Flights, seats and passengers by airport | Eurostat avia_tf_aca | 2025 | No | Observed |
| Airport-pair flights, seats and passengers | Eurostat avia_par_* | 2024 | No | Lagged |
| A220-specific activity | Eurostat aircraft taxonomy | 2025 | No | Not isolatable |
| A321XLR-specific activity | Eurostat aircraft taxonomy | 2025 | No | Not isolatable |
| Passenger itinerary and connecting airport | No complete public EU dataset | — | Potentially, with design | Commercial data required |
Sources: ACI/SEO, Eurostat and EUROCONTROL public documentation; reviewed 30 August 2026.
Act VII · The causal boundary
What about the A220 and A321XLR?
The mechanism is plausible: smaller, longer-range aircraft can lower the capacity threshold for a direct route. But Eurostat does not isolate the A220 cleanly, combines A321 variants, and public airport-pair data stop at 2024.
Verdict · cleared to publish
Europe is becoming less dependent on some traditional hubs, on some routes, through some airline models.
That is materially different from saying that hubs are obsolete.
- Flight activity recovered faster than connecting-network quality.
- Hub power shifted toward airports such as Istanbul and Athens while Frankfurt stayed below its 2019 position.
- Medium airports regained direct connectivity, but small airports did not.
- The aircraft mechanism is plausible, but not identified by the public data.
Methods · concise by design
Three evidence families. No blended denominator.
Eurostat counts commercial IFR movements, passenger flights, seats, and passengers. SEO/ACI NetScan measures weighted direct, indirect, total-air, and hub connectivity. Every 2025 recovery index compares one measure with its own 2019 level.