WP-017 established that sovereign bond spreads react to fiscal signals in under 24 hours but show no detectable response to structural capacity investment deficits over 76 months. The gap between these two reaction speeds implies a structural blind spot: financial markets cannot distinguish between capacity that is installed and capacity that is systemically integrated. This paper introduces the Integration Quality Score (IQS), a constructed index measuring the systemic integration quality of capacity investments across four dimensions: physical capacity (w=0.15), systemic linkage (w=0.40), policy anchor (w=0.30), and export potential (w=0.15). A three-case pilot produces the following scores: Finland onshore wind 4.1/10, Finland data centres 4.7/10, Denmark Ørsted offshore system 8.5/10. The 4.4-point gap between Finland wind and Denmark Ørsted is invisible in bond spread data — both countries show investment occurring. IQS is the energy-system analogue of HDCI (WP-016): a diagnostic instrument for what the market cannot see.
WP-017 demonstrated an asymmetric signal structure in sovereign credit markets: fiscal events produce sub-24-hour rating reactions, while structural capacity investment deficits produce no detectable reaction over 76 months. The conclusion was that bond spreads are necessary but insufficient indicators of long-term state sustainability — they measure the ability to service current debt, not the ability to build future capacity.
This paper addresses the next question: what would a sufficient indicator measure? The answer implicit in WP-017's cross-country typology (§3.6) is systemic integration quality — not whether investment has occurred, but whether investment is coupled to the broader system in a way that produces durable economic benefit. Finland has installed approximately 7,000 MW of onshore wind capacity. Denmark has built a comparable system around Ørsted and the energy islands. Bond spreads cannot distinguish between these two cases. IQS is designed to make this distinction measurable.
The structural parallel to WP-016 (HDCI) is precise. HDCI measures health data integration quality — the gap between data that exists in Finnish registries and care coordination that actually occurs. IQS measures energy investment integration quality — the gap between capacity that has been installed and systemic utilisation that produces economic value. Both instruments operationalise the same diagnostic principle: the system optimises for what it measures, and what it does not measure decays without sanction.
| Dimension | Weight | What it measures | Bond-visible? |
|---|---|---|---|
| D1 Physical capacity | 0.15 | Installed or committed MW/GW; capacity factor; grid connection status | Partially — investment flow appears in GDP |
| D2 Systemic linkage | 0.40 | Demand-response / flexibility framework (binding or voluntary); industrial coupling (PtX, waste heat, offtake); grid balancing participation; cross-sector integration | No — integration layer does not appear in fiscal aggregates |
| D3 Policy anchor | 0.30 | Legislative mandate; national strategy with measurable targets; state ownership or formal mandate; EU funding integration | No — policy quality not visible in spread |
| D4 Export potential | 0.15 | Commercial export pathway operational; international PPAs; technology export | Partially — export earnings appear in current account |
Formula: IQS = 0.15·D1 + 0.40·D2 + 0.30·D3 + 0.15·D4
Note (August 2026, pending): D2 as currently specified measures whether coupling exists (flexibility framework, industrial offtake, grid participation) but not who bears the resulting cost and who captures the resulting benefit. A candidate sub-indicator — provisionally "Fiscal/Systemic Capture" — would ask this directly: grid reinforcement cost, capacity opportunity cost, the system pressure created by inflexible load, public infrastructure cost, and net fiscal effect, set against who receives the investment's benefit. The Finland data centre case (§4.3, Heikura; Kangasharju & Pervilä) is a concrete instance of this question, not a general finding about data centres — it is one observation motivating the indicator, not evidence that scores it. Applying it consistently would require scoring all three pilot cases (including Bornholm's own grid-cost allocation) on the same basis, not only the case that motivated it. Marked empirical validation pending; not yet incorporated into D2 scoring or weights.
Cross-domain parallel (August 2026, pending): FSC as a general "distribution-gap" question, alongside WEM's endurance-gap. A PPA locks in a price for the investor but does not lock in the system's own endurance capacity: if firm, dispatchable capacity (e.g. CHP) is retiring while weather-dependent generation grows, a large continuous load can be contractually shielded from the price of the hours when wind does not carry the system — but the system still has to source real power for those hours regardless of what any individual PPA says. The PPA transfers price risk; it does not create the capacity that removes the underlying physical risk. This is the same structural pattern as the fiscal-capture question above, one level down: a headline aggregate (GDP, investment volume, MW installed) can improve while the distribution of who bears the resulting cost, and who captures the resulting benefit, moves in the opposite direction from what the aggregate suggests. WEM's endurance-gap and this FSC indicator are, on this reading, the same question asked in two different domains — physical system stress tolerance in one case, economic value distribution in the other — not two separate findings.
Macro context for the Heikura finding, verified against Bank of Finland forecasts (Talouskatsaus, June 2026; "Talous käänteen kynnyksellä"): Finnish GDP growth is forecast at +0.8% (2026), +1.6% (2027), +1.7% (2028); unemployment is forecast to remain at approximately 9.0% at the end of the forecast horizon; general government debt is forecast to rise to just under 97% of GDP by 2028; Q1 2026 GDP growth of 0.9% quarter-on-quarter was driven primarily by exports, with domestic demand remaining weak (Tilastokeskus). None of this establishes that data-centre-type investment specifically causes the gap between headline growth and domestic household benefit — that causal claim is not tested here. It establishes only that the forecast period in which large, capital-intensive, low-employment investment is arriving coincides with a forecast period of persistently high unemployment and rising public debt, which is the macro-level condition under which the distribution question above is not a hypothetical concern.
| Score | Level | Description |
|---|---|---|
| 0–2 | Absent | No evidence of integration; capacity isolated from system |
| 3–4 | Planned | Announced or under development; not operationalised or binding |
| 5–6 | Partial | Some elements operational; material gaps remain |
| 7–8 | Substantial | Operational with documented minor gaps |
| 9–10 | Full | Integrated, governed, commercially active across all indicators |
Approximately 5,500 MW operational at end-2025, with ~7,000 MW committed or under construction. Capacity factor 28–32%, consistent with Nordic conditions. Full grid connection; Finland's wind build is among the highest per-capita in the EU. D1 scores high — the physical investment is real and substantial.
This is the critical dimension. Finland has over 40 announced PtX projects with a combined investment potential of EUR 10 billion — but 2025 brought both postponements and cancellations. P2X Solutions' Harjavalta plant (small-scale) is operational. Tampere e-methane (Nordic Ren-Gas, 160 GWh capacity) targets 2026 commercial delivery. Plug Power announced 2.2 GW electrolyser plans (2023) with FID target 2025/2026 — status unclear.
What is absent: a national PtX roadmap mandating coupling between wind capacity and industrial demand. TEM's 800 MW flexibility mechanism is voluntary and market-based (DT-004). No binding demand-response obligation for wind producers. Curtailment already occurring on high-wind days — the integration layer that would absorb excess generation is planned but not operationalised at system scale. Finland imports wind technology (Vestas, Siemens Gamesa); no domestic technology export base.
Finland is the only Nordic country without a capacity mechanism (DT-001). No national PtX target. Permitting has been streamlined (fast-track for green hydrogen added January 2025), and TEM has appointed a working group preparing hydrogen market legislation (due August 2026). These are partial signals: procedural progress without a binding integration mandate.
No operational wind electricity export strategy. Hydrogen export ambitions exist (Kokkola liquid hydrogen for Port export, Oulu e-fuels) but all are pre-FID. Finland imports wind turbine technology; no technology export pathway is operational.
IQS = 0.15·8 + 0.40·3 + 0.30·4 + 0.15·3 = 4.1 / 10
High physical capacity (D1=8) combined with absent systemic integration (D2=3) and weak policy anchor (D3=4) is the defining signature of Type III pipeline activation: permissive attraction without integration.
Approximately 500 MW operational at end-2025. Pipeline includes Google, Microsoft, Meta, and DayOne (560 MW Klaukkala announced Q1 2026). However, Google Muhos (multi-billion EUR) paused October 2025 following electricity tax signals — demonstrating pipeline fragility. D1 scores moderate: substantial operational capacity, but pipeline volatility reduces certainty.
Finland's data centre waste heat integration is a genuine partial success. Google Hamina (operational end-2025): covers ~80% of local district heating demand. Microsoft Espoo (AFRY-designed, 350 MW thermal, Fortum EUR 225M investment): world's largest data centre heat recovery project, heating ~100,000 homes from 2025–2026. These are real, operational integrations — not pilot projects.
What remains absent: binding demand-response obligation. Data centres can connect via PPA without creating new capacity (DT-004). Grid balancing participation is voluntary. Fingrid paused new southern connections pending grid reinforcement — the system is approaching its absorption limit for flat 24/7 load without mandatory flexibility. Germany's EnEfG mandates 10% waste heat utilisation from July 2026, rising to 20% by 2028; Finland has no equivalent obligation.
PM Orpo appointed rapporteur June 2025 for national DC roadmap. But the dominant policy signal of 2025 was reactive: electricity tax shift (HE 156/2025 vp, category II → I from July 2026) prompted Google to pause Muhos. A new support scheme has since moved from concept to concrete design (Ministry of Economic Affairs and Employment, December 2025 – March 2026): a fixed-term, ten-year electricity-tax-based refund, capped at approximately EUR 30 million per year in forgone revenue, conditional on data centre registration and either waste-heat utilisation or an energy-efficiency standard exceeding existing electricity-tax-relief requirements. As of this writing the scheme remains in preparation, not enacted. A national DC strategy with binding integration targets does not yet exist. D3 reflects a rapporteur mandate and a support scheme in preparation, set against the absence of any binding strategy or targets — a formal process exists, but nothing yet binds.
May–August 2026 — fiscal/systemic capture evidence (see D2 note, §2.1, and §5 methodology): not scored here. Verohallinto (Finnish Tax Administration) Director General Markku Heikura made public statements, based on registry material rather than published registry data itself, disclosing the following figures (Hufvudstadsbladet, May 2026; recirculated on LinkedIn): the largest single tax revenue observed from one data centre is €11 million; the largest single VAT refund paid to a data centre is €240 million — independently corroborated by Finnish Wikipedia's "Datakeskukset Suomessa" summary of the same HBL interview, and discussed on Finnish industry forums citing the same figure. Of Finland's 44 registered data centres, only seven employ more than 20 people; combined, the 44 centres produced approximately €7.7 million in tax revenue in 2024. Heikura's conclusion: "In some cases, tax revenues can even be said to be negative." The mechanism is structural — intra-group transfer pricing, sales booked outside Finland, investment depreciation allowances, and VAT refunds on capital expenditure combine to produce net-negative fiscal integration at present. Heikura explicitly noted that individual companies operate within existing law; the structural problem is the model, not compliance. The data centre industry association's dispute of Heikura's characterisation, referenced imprecisely in an earlier version of this note, is now identifiable: EK (Confederation of Finnish Industries) and FDCA (Finnish Data Center Association) commissioned Ramboll (published September 2025) for a projected-impact study, based on an announced investment base of €12bn (potential up to €30bn): construction-phase tax impact €1.7bn (property, corporate, income, and VAT combined), production-phase employment 9,900 person-years and tax impact €400M/year from 2030 onward. In May 2026, EK issued a follow-up statement ("Datakeskusten talousvaikutuksista tarvitaan faktoihin perustuvaa keskusteluun"), asserting that public discussion had included incorrect claims — understood here as EK's response to the Heikura disclosure above.
Floor and ceiling, not contradiction (August 2026). Heikura measures realised state (2024, 44 operating centres, €7.7M actual combined tax revenue) — a floor. EK/Ramboll measures projected state (2030, if the announced €12–30bn investment pipeline materialises) — a ceiling. These are not in conflict on their face: both can be true simultaneously if Fingrid's own connection-agreement pipeline (§ note elsewhere, ACI WEM instrument, August 2026 — near 5GW of data-centre connection agreements, Fingrid's own estimate of at least five years for full realisation) proceeds as projected. This is the same diverted-versus-induced-demand distinction used throughout ACI's CN-030: a realised-now figure and a projected-if-realised figure measure different temporal claims, and disagreement between them is not evidence that either party is wrong. Which figure turns out closer to 2030 outcomes is an empirical question this note does not adjudicate and cannot yet answer.
Independent corroboration: University of Helsinki, "Datakeskusten faktat ja fiktiot" (Kangasharju & Pervilä, 2026). Jussi Kangasharju (professor of computer science) and Mikko Pervilä (visiting researcher) reach the same structural conclusion by an independent route — academic analysis of the VAT mechanism itself, not tax registry disclosure. Their finding: VAT refunds to data centres likely exceed the combined total of corporate tax and property tax that the same centres pay. Data centres claim substantial VAT deductions on capital purchases, but because they do not sell VAT-liable services within Finland, the deduction — designed as a cost-neutral accounting pass-through for firms with matching taxable sales — has no offsetting VAT collection to net against, and functions instead as a direct fiscal transfer. The report separately flags that publicly cited power-demand and employment figures for planned data centres are routinely inflated relative to what is realised.
Why this is not a D3 revision. D3 measures whether the state can steer the system — legislative mandate, national strategy, ownership, EU funding integration, supply independence. Fiscal capture measures whether the system produces a net linkback to the public purse. These are different questions; conflating them was an earlier error in this note (corrected August 2026 — an intermediate version of this paper scored D3 at 3/10 citing the Heikura finding, which has been reverted). Two independent sources reaching the same fiscal-capture conclusion by different methods within three months of each other is a strong empirical signal — but it is evidence for the pending Fiscal/Systemic Capture indicator (§2.1), not for D3. It is recorded here, adjacent to D3, because a reader assessing "policy anchor" will want it in view — not because it is counted in the D3 score.
Finland is positioning as a sustainable data hub (Bloomberg: "redefining sustainable digital economy"). The waste heat model has reputational export potential. Data services are inherently non-exportable as physical product, but Finland's model may influence EU data centre regulation — a soft export of governance approach.
IQS = 0.15·6 + 0.40·5 + 0.30·4 + 0.15·4 = 4.7 / 10 (corrected August 2026: a May 2026 revision had lowered D3 to 3/10 citing fiscal-capture evidence that does not fall under D3's definition; reverted to 4/10. The fiscal-capture finding itself stands, independently corroborated by two sources, but is recorded against the pending FSC indicator, §2.1, not against D3.)
Finland's data centre case is structurally more integrated than its wind case — the waste heat linkage is operational and at scale. But the absence of binding flexibility obligations and the reactive policy environment keep D2 and D3 in the partial-to-low range. Separately, and on a different axis, the confirmed net-negative fiscal capture (Heikura May 2026; Kangasharju & Pervilä August 2026) is exactly the kind of finding the pending FSC indicator exists to score once validated across all three pilot cases. The Google Muhos pause illustrates the fragility of Type III pipeline activation: policy ambiguity converts investment pipeline into investment withdrawal risk.
Scope note (corrected August 2026): an earlier version of this section scored a single "Ørsted + Energy Islands" unit, attributing Bornholm Energy Island's capacity to Ørsted's own D1 and folding UK/Dutch/German corporate assets into what read as a Danish national figure. Bornholm is not Ørsted's: it is owned and delivered by Energinet (Danish state TSO) and 50Hertz (German TSO); the December 2024 tender for its offshore wind generation capacity attracted zero bids. This version separates the two units explicitly wherever they diverge: Ørsted A/S (the corporate entity, 50.1% Danish state-owned) and the Danish offshore wind system (Ørsted plus Energinet/50Hertz-run grid infrastructure plus other developers). D1 and D4 below are scored for the system, since export and pipeline scale are genuinely system-level; D3's ownership fact is Ørsted-specific and marked as such. Where this changes the substance of a claim, it is noted inline.
Ørsted itself has 7.5 GW installed across Denmark, UK, Netherlands, and Germany, with 2.5 GW under construction — 10.2 GW installed offshore capacity globally as of late 2025, 8.1 GW under construction. This is verified, operating capacity. Bornholm Energy Island (3 GW planned, HVDC to Danish and German grids, EU CEF grant EUR 645M signed September 2025) is not Ørsted's asset — it is delivered by Energinet and 50Hertz, and the December 2024 tender for the wind generation capacity itself attracted zero bids, pushing the government toward contracts-for-difference subsidy instead. Onshore infrastructure alone (converter buildings) has already run 85% over budget (DKK 1.7bn overrun disclosed 2026). North Sea Energy Island (3–10 GW planned) and Ørsted+CIP's 5.2 GW open-door projects remain pipeline, not installed capacity. D1 is scored down from the prior version to reflect that a material share of the previously cited "physical capacity" is planned system infrastructure with an unresolved procurement failure at its centre, not built or contracted generation.
Bornholm Energy Island is a genuine systemic-linkage architecture: the first cross-border hybrid offshore wind project, jointly financed by two countries, connecting to two national grids — this remains true regardless of which entity delivers it. Energinet (state TSO, not Ørsted) has the formal mandate to integrate offshore wind into grid architecture, and Denmark's near-term grid is designed around wind variability rather than retrofitted for it.
PtX integration: Denmark has modelled offshore electrolysis with HVDC for distances under 350 km, with onshore electrolysis waste heat reducing LCOH by 10–30%. Export scenarios modelled: 2 Mt hydrogen/year to Germany by 2045, e-SAF at EUR 170/MWh. World Hydrogen Week 2025 hosted in Copenhagen. D2 scores 8 rather than 10 because full commercial operationalisation of hydrogen export is still in development.
The Danish state owns 50.1% of Ørsted A/S — an Ørsted-specific fact, not a system-wide one. Separately, at system level: parliamentary majority for energy islands, 171/179 MPs in 2020, the broadest cross-party consensus in Danish energy history; EU PCI (Project of Common Interest) status confers regulatory priority; Danish wind policy continuity since 1979 (Vestas V10 model) — 46 years of consistent industrial policy direction. D3 scores 9: state ownership of the anchor actor, parliamentary consensus, EU backing, and historical continuity combine into the most robust policy anchor in the pilot set — this dimension is not materially affected by the Ørsted/system scope correction above, since ownership, parliamentary mandate, and EU status are all genuinely verified at their respective levels.
Ørsted is the world's largest offshore wind developer — technology export is embedded in the business model. Vestas (Danish) is a top-5 global turbine manufacturer. Bornholm Energy Island is explicitly positioned as an EU blueprint for future cross-border offshore projects, independent of which entity constructs it. Hydrogen export: commercial modelling active (2 Mt/year Germany). e-SAF: pathway commercially defined above EUR 170/MWh. PPAs operational across multiple markets. D4 scores 9: commercial export is operational, not planned.
Ørsted's high D1/D3/D4 scores describe integration architecture, not financial health, and the two have diverged sharply. In August 2025 Ørsted announced a DKK 60 billion rights issue, later completed at approximately DKK 59.6 billion; the Danish state subscribed its full 50.1% pro rata share (451.5 million shares). The issue followed the cancellation of a planned partial sale of the US Sunrise Wind project, deemed unviable under US regulatory conditions. Separately, Ørsted divested 50% of Hornsea 3 to Apollo-managed funds (agreement November 2025, completed December 2025, ~DKK 39bn value) and announced a reduction of roughly 2,000 positions — from ~8,000 to ~6,000 employees — by the end of 2027. Nine-month 2025 return on capital employed was reported at approximately 2.0%. None of this appears in the scoring above, because none of it is a D1–D4 fact as specified — it is a capture/distribution fact: the entity with the highest IQS score in this pilot set was simultaneously undergoing a state-backed recapitalisation, asset sales, and workforce reduction. This is treated in this note as motivating evidence for the pending Fiscal/Systemic Capture indicator (§2.1, §8.2) — the clearest available instance in the pilot set of what that indicator is designed to check for: high measured integration co-occurring with financial fragility that a capacity-and-integration score alone does not detect.
IQS = 0.15·8 + 0.40·8 + 0.30·9 + 0.15·9 = 8.5 / 10 (corrected August 2026: D1 9→8 following the Ørsted/system scope correction above; previously 8.6)
Denmark's offshore wind system, with Ørsted as its anchor actor, remains the reference case for Type I pipeline activation: state ownership, parliamentary mandate, EU-level system architecture, and operational export pathways. The IQS gap between Finland wind and Denmark is invisible in bond spread data. Both countries show investment occurring; FI-DK spread widens in 2025 for reasons entirely unrelated to energy integration quality (Novo Nordisk corporate shock, §2.4 in WP-017). Separately, and just as invisible to bond spreads: the anchor actor's own capital position is under real strain at the same time its IQS score is the highest in the pilot set — see the capital-position finding above.
| Case | D1 Physical | D2 Linkage | D3 Policy | D4 Export | IQS | Type |
|---|---|---|---|---|---|---|
| Finland wind | 8 | 3 | 4 | 3 | 4.1 | III — permissive |
| Finland DC | 6 | 5 | 4 | 4 | 4.7 | III — partial linkage |
| DK Ørsted | 8 | 8 | 9 | 9 | 8.5 | I — national champion |
The integration gap between Finland wind (IQS 4.1) and Denmark Ørsted (IQS 8.5) is 4.4 points on a 10-point scale — a structurally significant difference. Both cases show investment occurring. Bond spreads cannot detect this gap because the integration layer (D2: systemic linkage, D3: policy anchor) does not appear in fiscal aggregates, current account flows, or the five-pillar rating model inputs.
Finland's data centre case (IQS 4.7) occupies a middle position: the waste heat integration is operationally real and distinguishes it from the wind case. But the absence of binding flexibility obligations and policy fragility (Google Muhos pause) confirm that Type III pipeline activation — even when partially integrated — is structurally less robust than Type I national champion architecture.
The finding operationalises the Salazar mechanism hypothesis from WP-017: rating pressure rewards front-loaded fiscal consolidation while remaining blind to integration quality. A government that invests in Type III pipeline activation (permissive attraction, no integration mandate) improves its short-term investment flow appearance identically to a government investing in Type I (national champion with full integration). The market cannot distinguish them; the governance implication is that the political pressure to move from Type III to Type I is systematically absent.
The structural parallel between IQS and HDCI (WP-016) is the organising insight of this paper. Both instruments operationalise the same diagnostic principle across different sectors:
| HDCI (WP-016) | IQS (WP-018) | |
|---|---|---|
| Sector | Health data infrastructure | Energy capacity investment |
| What it measures | Integration quality of care data → care coordination | Integration quality of capacity → systemic utilisation |
| What existing metrics measure | Data volume, cost, queue length | Investment MW, GDP contribution, fiscal balance |
| Bond/market visibility | Not visible — integration failure is fiscal-neutral short-term | Not visible — integration quality does not appear in spread |
| D-suppression mechanism | No integration measurement → no sanction → care pathways fragment | No integration measurement → no sanction → capacity unintegrated |
| Pilot | DT-006 Pohjois-Savo (planned) | Three-case pilot (this paper) |
The shared structure is: measurement gap → sanction gap → correction gap → structural decay. HDCI and IQS are both instruments for closing the measurement gap. They cannot, by themselves, create the sanction — but without measurement, sanction is structurally impossible.
IQS v0.1 has five active limitations. First, weights are theoretical priors: D2 (0.40) is assigned the highest weight on the basis that systemic linkage is the least-visible and most policy-relevant dimension, but this has not been empirically validated. Second, scoring is analyst-dependent: inter-rater reliability has not been tested. Third, the pilot is three cases: generalisation to OECD-wide typology requires a larger, systematically selected case set. Fourth, predictive validity is undemonstrated: IQS must be tested against observable outcomes — energy system stress events, industrial energy costs, export earnings — before the instrument can be used for policy recommendations. Fifth, the D1 dimension is not sign-stable across applications: in the three pilot cases D1 measures installed capacity (high = more capacity, treated as better-integrated potential); in the §8.3 maintenance-deficit application D1 is redefined as deficit magnitude (high = larger shortfall, worse). The two are not the same variable and the resulting composites are not on a common scale — §8.3's result is reported as IQS-M, not IQS, for this reason.
| Step | Description | Horizon |
|---|---|---|
| Inter-rater reliability | Score all three cases with second analyst; measure Cohen's κ per dimension | Immediate |
| Weight sensitivity | Recalculate IQS under alternative weight vectors; report rank stability | Immediate |
| Case expansion | Add Spain (RRF absorption), Poland (RRF + coal transition), Norway (hydro integration), Germany (Energiewende) | Next version |
| Outcome validation | Correlate IQS scores with energy price volatility, industrial competitiveness, export earnings across OECD | v1.0 target |
| IQS × bond spread | Test whether IQS adds explanatory power to spread models for long-horizon stress prediction | v1.0 target |
| FSC definition and cross-case test | Formalise the Fiscal/Systemic Capture indicator (§2.1) precisely, then apply the same test to all three pilot cases — including the Danish system's own grid-cost allocation on Bornholm (Energinet/50Hertz, not Ørsted) and Ørsted's own capital position (§5, finding box) — before any score is assigned. Not: rescore Finland DC on the basis of the Heikura/Kangasharju-Pervilä finding alone. | Next version |
Note (August 2026): the IQS framework as specified answers "how much capacity, and how well is it integrated" (D1→D2→D3→D4). The pending FSC indicator points at a third, distinct question the current framework does not ask: once capacity exists and is integrated, who bears the cost, who captures the benefit, and does net value remain in the system? Capacity formation is comparatively easy to observe — GDP, investment flow, MW installed all register it. Integration is harder to see, which is IQS's own founding premise (WP-017's finding that bond spreads miss it entirely). Capture — the distributional question — may be harder still, and by a different route: not absent from official statistics because no one measures it, but because standard fiscal aggregates net flows that move in opposite directions (tax paid, VAT refunded) into a single line that looks unremarkable. If this pattern holds across cases beyond the one that surfaced it, it would describe a general property of how these systems are measured, not a Finland-specific or data-centre-specific finding. That generalisation claim is exactly what the cross-case test above exists to check — it is not yet established, and should not be treated as such until it is.
Construct note (August 2026): in the main IQS framework, a high D1 score means substantial capacity exists — more is better. Below, D1 is redefined as deficit size, where a high score means the shortfall is large — more is worse. This is not the same variable with a different label; the sign is inverted. The resulting composite is therefore not comparable to the three pilot cases (§3–§5) on the same 0–10 scale, and is renamed IQS-M to make that explicit rather than presented as a fourth IQS score alongside 4.1/4.7/8.5. This limitation is not currently listed in §8.1 and is added there as a fifth item.
The IQS-M diagnostic logic is nonetheless useful on its own terms: the same four-part question (physical need, integration into a correction mechanism, policy mandate, economic utilisation) applies to the maintenance and renewal of existing infrastructure — a domain where the gap between documented need and funded correction is equally large and equally invisible to bond market pricing.
Finland's publicly documented infrastructure maintenance deficit provides a test case anchored in official sources:
| Asset class | Documented deficit | Source | Trajectory |
|---|---|---|---|
| State transport network (roads, railways, waterways) | €4.2 billion (end-2024) | TAE 2026 (budjetti.vm.fi) — official, budget-line figure | Growing +€52M/year in 2024; projected to exceed €5 billion by 2028 at current funding levels (Väylävirasto) |
| Municipal water supply and sewage networks | No official aggregate figure | Vesilaitosyhdistys 2025; vesihuoltolaki reform 2026 acknowledges problem | Widely recognised as growing; new Water Services Act (2026) mandates long-term investment planning |
| Built environment — infrastructure total (ROTI 2023) | €41.6 billion (utility infrastructure: water, energy, telecoms, waste) | ROTI 2023 independent expert assessment (Green Building Council Finland) | Grown from €24.9B (2000) to €77.5B (2021) for entire building stock; utility infrastructure assessed separately |
The policy context compounds the deficit. State transport network base funding is set at approximately €1.4 billion per year for 2025–2028. Väylävirasto projects that funding will fall to an "extremely tight level" in 2027–2029 — coinciding precisely with the energy capacity intervention window identified in SM-006. SKAL estimates that a permanent additional €300 million per year would be required simply to stop the deficit from growing; the government's €520 million three-year package (2024–2026) is insufficient and expires before 2027.
Preliminary IQS-M scoring — infrastructure maintenance deficit (not comparable to IQS 0–10 pilot scores above; see construct note):
| Dimension | Score | Rationale |
|---|---|---|
| D1-M Deficit magnitude (inverted: high = large shortfall) | 8 | Deficit is large, documented, and growing — €4.2B state network alone. Physical need is well-quantified. |
| D2 Systemic linkage (integration into correction programme) | 3 | Correction programme exists but is underfunded. Base funding 2027–2029 falls below maintenance threshold. Correction is planned on paper, not operationally resourced. |
| D3 Policy anchor (mandate and funding) | 3 | Väylävirasto identifies need; Liikenne 12 plan acknowledges shortfall. But concrete funding commitment for 2027+ is absent — government programme expires 2026, no successor programme confirmed. |
| D4 Export/economic return potential | 2 | Infrastructure maintenance has no export pathway. Economic return is internal (competitiveness, logistics cost) but not separately measured or monetised as investment return. |
Preliminary IQS-M = 0.15·8 + 0.40·3 + 0.30·3 + 0.15·2 = 3.6/10 (corrected arithmetic, August 2026 — was misstated as 3.3; IQS-M scale, not comparable to IQS 4.1/4.7/8.5 above)
This score is lower than Finland's wind IQS (4.1) and data centre IQS (4.7) — reflecting that infrastructure maintenance has a better-documented physical need (D1) but weaker systemic integration and policy anchor than new capacity investment. The pattern is structurally identical: a large, growing, officially acknowledged need without a funded correction mechanism that closes the gap within the relevant time horizon.
Finland's €4.2 billion documented state infrastructure maintenance deficit (TAE 2026) exhibits the same IQS pattern as the energy investment cases: high physical need (D1), weak systemic correction mechanism (D2), insufficient policy anchor to close the gap (D3), no economic return pathway (D4). All four investment pressure streams — energy capacity, data centre integration, infrastructure maintenance, and defence — compete for the same constrained public investment capacity over the 2027–2030 horizon. Bond spreads are equally blind to the maintenance deficit as to the energy capacity deficit: both are fiscal-neutral in the short term until they produce service failures or acute cost events. This is the generalised form of WP-017's structural invisibility finding applied to a different asset class.
Methodological note: the preliminary IQS score for infrastructure maintenance is an illustrative extension, not a validated case study. The D4 dimension (export potential) may not be the appropriate indicator for maintenance infrastructure — a domain-adjusted framework may be required for non-tradeable public goods.