European Energy Investment: A Category I / Category II Proxy Analysis

RQM-001 requires, as evidence condition E-1, at least two OECD country pairs showing differential Category I / Category II allocation ratios correlated with capital architecture type. Before that cross-country comparison can be constructed, a prior step is needed: establishing whether the Category I / Category II distinction is empirically observable at all in available energy investment data, and whether the proxy indicators are coherent enough to support the cross-country analysis E-1 requires.

This assessment addresses that prior step for the European energy system, using IEA investment and capacity data as the primary source. It does not claim to resolve E-1 — that requires country-pair data and a capital architecture control variable. It establishes whether the diagnostic partition is visible in the aggregate data, and identifies which specific data series would need to be disaggregated to conduct the full E-1 test.

The assessment is explicitly a proxy analysis. The Category I / Category II distinction was developed in WP-008 for infrastructure capital allocation broadly — its application to energy investment data requires definitional choices that are made transparent here and are open to revision.

In WP-008, Category I (consumption-binding) infrastructure commits system capacity to a fixed load profile, while Category II (stability-providing) infrastructure increases system flexibility and recovery capacity. Applying this distinction to energy investment requires translation, since energy sector investment categories do not map cleanly onto WP-008's original typology.

For this proxy analysis, the translation is as follows. Category I covers investment that primarily increases electricity supply capacity without providing dispatchable control — variable renewable generation (solar PV, wind) and consumption-binding load infrastructure (hyperscale datacenters, large industrial electrification). Category II covers investment that primarily increases the system's ability to govern its own operation under stress — grid reinforcement, dispatchable and flexible generation, energy storage, and demand-side flexibility infrastructure.

One deliberate ambiguity: grid investment sits at the boundary. Grid reinforcement enables both renewable integration (Category I logic) and system resilience (Category II logic). For this analysis, grid investment is treated as Category II on the grounds that it conditions the system's capacity to absorb stress, regardless of the generation mix it serves.

IEA data for the European Union shows a large and widening asymmetry between investment in variable renewable generation and investment in the flexibility and stability infrastructure that conditions how that generation can be used. The asymmetry has been accelerating.

The diagnostic signal here is not the absolute level of either category but the trajectory divergence. Renewable generation investment grew 35-fold relative to fossil fuels over a decade. Grid investment was essentially flat from 2015 to 2022 at approximately USD 300 billion globally per year before beginning to recover. The gap between what the system generates and what the system can govern is widening — and has been widening for most of the period during which the current infrastructure commitment was being made.

If the Category I / Category II imbalance is real, it should produce observable operational signals in the electricity system — not only in investment data. Three signal types are available from IEA data for 2024–2025.

Signal S-2 is particularly diagnostic for the RQM-001 framework. The EU electricity system in H1 2025 experienced a compound stress event: two major variable sources simultaneously underperformed. The system response was to draw on dispatchable fossil generation — assets that are not growing and whose investment share is at 1:35 relative to renewables. This is a direct observation of Category II dependency under stress, using exactly the legacy reserve assets that WP-001's Black Period analysis identifies as the critical endurance layer.

RQM-001 predicts that countries with different capital architectures will show different Category I / Category II allocation patterns. The following comparison is based on structural capital architecture data, not yet on disaggregated Category I / Category II investment ratios — that is the E-1 test that DA-004 is scoping. The structural patterns are consistent with the prediction.

The structural patterns are consistent with RQM-001's prediction, but this is not yet an E-1 test. E-1 requires disaggregated investment ratio data showing that Norway and France have measurably higher Category II shares than Finland and Germany, controlled for sectoral mix. That data is not yet assembled. DA-004 identifies the structural plausibility of the prediction; the empirical test remains to be constructed.

DA-004 establishes three things. First, the Category I / Category II distinction is observably present in European energy investment data — the 35:1 renewable-to-fossil ratio, the near-stagnation of grid investment 2015–2022, and the documented grid connection queues and curtailment events are all consistent with a system where stability-providing investment has not grown proportionally to consumption-enabling commitment.

Second, the H1 2025 EU electricity stress event is a direct empirical observation of Category II dependency under compound meteorological stress — the system revealed its continued dependence on legacy dispatchable assets whose investment share is minimal. This is not a crisis, but it is a palpable signal of the endurance structure WP-001 identifies as the diagnostic layer.

Third, the capital architecture comparison is structurally consistent with RQM-001's prediction, but is not yet an empirical test of it. The prediction is that Norway and France would show higher Category II investment shares than Finland and Germany when controlled for sectoral mix — this is testable with disaggregated IEA or national energy investment data, but that data assembly has not been done here.

To complete the E-1 test as specified in RQM-001, three specific data series are required beyond what is available in public IEA aggregate statistics.

First, disaggregated national energy investment data by generation type and grid/flexibility category for Norway, France, Finland, and Germany — ideally 2015–2024 to capture the full trajectory divergence period. This is available in principle from IEA's World Energy Investment country annexes and from national energy agency statistics (Fingrid, Statnett, RTE, Bundesnetzagentur).

Second, pension fund and infrastructure fund allocation data by sub-asset class showing the share going to variable renewable generation versus dispatchable/storage/grid assets. OECD's Long-Term Investing of Large Pension Funds survey collects infrastructure sub-sector data but the 2023 edition does not disaggregate at Category I / Category II resolution. Direct fund disclosure data or Preqin infrastructure database access would be required.

Third, for the capital architecture control variable, national statistics on the share of infrastructure investment financed through public-mandate versus private-market vehicles. This is partially derivable from existing OECD data but would require country-level assembly rather than aggregate reporting.

These data requirements are not prohibitive for a working paper. They are more demanding than can be assembled from publicly available aggregate statistics alone, and would benefit from access to Preqin or similar private-market data infrastructure.