Methodology
How an asset earns its score, how nations are ranked from those scores, how confidence is assessed separately from importance, and what this design does badly. The rubric follows the GINC National Infrastructure Index methodology, taxonomy version 1.0.
The unit of analysis
The unit is the strategic infrastructure asset-system, not the building. A port, a power station, a refinery, a dam, a data centre, a mine, a pipeline, a railway corridor, a transmission network, a port cluster, a satellite constellation and a national payment system are all legitimate single units. What decides the boundary is where the strategic function actually lives: the Trans-Siberian Railway is one asset because its significance is the corridor, while a substation on a national grid is not, because its significance is the grid.
This is also where the register is weakest. Review found 109 records that bundle two genuinely separate assets into one line - a launch complex with the base beside it, two coal terminals at one bay. They are flagged and need splitting.
The five dimensions
Each asset is scored nought to five on five dimensions. They are deliberately different questions, and an asset can be extreme on one and negligible on another: a national museum scores high on throughput and near zero on flow criticality, which is exactly what separates it from a grid.
Flow Criticality
What breaks
How consequential disruption would be to nationally important flows: electricity, fuel, water, food, trade, logistics, data, communications, finance, industrial inputs, defence capability, public services.
| Score | What it means |
|---|
| 0 | Negligible national consequence |
| 1 | Primarily local importance |
| 2 | Important to a city, region or niche sector |
| 3 | Significant national or major-sector importance |
| 4 | Critical to a major national system |
| 5 | Foundational national or international chokepoint |
Geographic Reach
How far it carries
The geographic scale over which the asset's effects are felt. This is reach, not size: a single small cable landing station can be international.
| Score | What it means |
|---|
| 0 | Site only |
| 1 | Local |
| 2 | Regional |
| 3 | Multi-region or national sector |
| 4 | National |
| 5 | International, continental or global |
Systemic Centrality
Where it sits in the system
The asset's role as hub, backbone, gateway or concentration point. Judged on share of national throughput, network centrality, gateway position, concentration of routes, dependency of downstream systems, and whether it acts as a control point.
| Score | What it means |
|---|
| 0 | Peripheral |
| 1 | Minor node |
| 2 | Important node |
| 3 | Major hub |
| 4 | National backbone or gateway |
| 5 | Dominant chokepoint or system-control node |
Low Substitutability
How hard to replace
Difficulty of replacing, bypassing or replicating the asset: alternative routes, spare capacity, construction lead time, technical uniqueness, specialist workforce, geological or geographic uniqueness, permitting, technology dependencies, capital cost. The name carries its direction - 5 means hard to substitute.
| Score | What it means |
|---|
| 0 | Easily substituted |
| 1 | Multiple immediate alternatives |
| 2 | Alternatives available with limited disruption |
| 3 | Replacement is difficult or slow |
| 4 | Very limited substitutes |
| 5 | Unique or effectively irreplaceable in the medium term |
Economic Throughput
How much runs through it
The scale of activity enabled or concentrated: tonnes, passengers, megawatts, barrels, gas volume, water supplied, data traffic, transaction value, manufacturing output, commodity value, population served, GDP exposure, defence capability enabled.
| Score | What it means |
|---|
| 0 | Minimal measurable throughput |
| 1 | Small |
| 2 | Moderate |
| 3 | Large |
| 4 | Very large |
| 5 | Exceptional national or global scale |
The strategic score
strategicScore = flowCriticality + geographicReach + systemicCentrality + lowSubstitutability + economicThroughput
range 0 to 25, all five dimensions equally weighted
Equal weighting is a first-generation choice, not a finding. Later versions may calibrate weights empirically against productivity, trade competitiveness, disaster recovery or observed disruption. Until that testing exists, equal weights are the honest default because they encode no untested theory about which dimension matters most.
Scores are absolute and comparable across nations and sub-sectors. They are not normalised within a list. A mid-ranked seaport in a trading state legitimately outscores the best asset in that nation’s solid-waste sub-sector, and that is the whole point: it lets a national top 100 emerge from competition rather than from quotas.
Ranking
A nation’s top 100 is generated by letting every scored candidate in that country compete on strategic score, with no fixed sector quotas, so the country’s real infrastructure structure is what shows. Records that review judged not to be discrete assets are excluded. Ties break in this order:
1. Systemic Centrality 2. Low Substitutability 3. Flow Criticality
4. Economic Throughput 5. Geographic Reach 6. Evidence Confidence
Evidence confidence
Confidence answers a different question from score: how well is this specific named asset documented? It is not a discount on importance. A closed state’s principal enrichment plant is a high score and a medium confidence, and recording it any other way would penalise opacity by pretending the asset matters less than it does.
| Band | Meaning |
|---|
| High | 0.90 to 1.00. Real, well documented, correctly named, with an article about the asset itself or equivalent public documentation. |
| Medium | 0.70 to 0.89. Certainly real and correctly identified, but documented only inside a broader article, or thinly covered in English. |
| Low | 0.50 to 0.69. Believed real, but the specific named asset could not be pinned down, or the naming is uncertain. |
| Provisional | Below 0.50. Existence as described could not be established. |
The single largest correction this register has undergone was to confidence, not score. The assessment pass rated 86 per cent of top-100 assets high confidence; individual review put that at 37 per cent. It had been rating how sure it was the asset mattered rather than how well it is documented.
Classification rules
An asset takes one primary capability group and one primary sub-sector, chosen by its dominant strategic function. Cross-sector dependence is captured as a secondary relationship rather than by listing the asset twice. The rules that settle the recurring ambiguities:
A gas field
Resource Infrastructure, Oil and Gas Extraction - energy commodity production is extraction.
An oil refinery
Resource Infrastructure, Processing and Refining - not electricity infrastructure.
A gas-fired power station
Energy Infrastructure, Thermal Generation - it converts fuel to electricity.
A hydroelectric dam
Energy Infrastructure, Hydropower - the water function is secondary.
A multi-purpose reservoir
Water, Dams Reservoirs and Flood Defence - unless generation dominates.
A nuclear power plant
Energy Infrastructure, Nuclear Generation - civil energy.
A nuclear weapons complex
Space, Aerospace and Defence - the defence function dominates.
A military shipyard
Space, Aerospace and Defence, Manufacturing Facilities - commercial yards go to Advanced Manufacturing.
A securities depository
Financial Infrastructure, Markets Exchanges and Depositories.
A government complex
Social Infrastructure, Government and Public Administration.
A cable landing station
Communications and Digital, Subsea Cables and International Gateways.
A grain terminal
Agriculture and Food, Storage Cold Chain and Markets - the port link is secondary.
A mining export railway
Transport, Rail Infrastructure - the mine link is secondary.
How the register was built, and what that costs
Candidate discovery ran sub-sector first rather than from a hand-curated national list: for each nation, every one of the 50 sub-sectors was searched for its leading assets, and the national top 100 was then generated from the merged pool. This avoids the fame bias of curating a top 100 directly, and it is why obscure but critical assets appear above famous ones.
It also introduces the register’s main known flaw. Taking a fixed number of assets per sub-sector is an implicit equal quota one level down from the one the methodology warns against. Groups with six sub-sectors contribute far more candidates than Financial Infrastructure with three, while a genuinely deep sub-sector - seaports in a trading state, oil and gas in a petro-state - is truncated. The derived national top 100 therefore over-represents thin sub-sectors, and capability-group shares computed from it run flatter than reality. The fix is deeper discovery in thick sub-sectors, not a change to the rubric.
Beyond the twenty. Every other nation began with a list of its fifty most strategic assets, scored in a second pass against the same rubric with a per‑nation economic‑scale tier supplied so that a small state's sole port and a large state's fortieth port land on one absolute scale. A national expansion then searched each of those nations for up to 150 further assets, capability group by capability group, with the nation's existing list supplied as the calibration anchor and an explicit instruction to stop short rather than pad. It added 22,670 assets. Ninety‑nine nations were searched to the target; seventy‑eight stopped at a ceiling, most of them microstates and small island economies that do not contain two hundred strategic assets. Most of what the expansion found is each nation's second tier, and it scores that way: a mean of 12.0 against 14.7 for the top‑50 lists.
Deduplication and review
Duplicates were removed in two rounds. The first matched names within each country on token and character similarity, merging close matches automatically and sending 837 borderline pairs to individual adjudication. The second round re-ran the match after review had corrected the names, on the reasoning that a duplicate written very differently becomes visible once both records carry their proper name. 1,022 records were merged away in those rounds. Two further rounds then treated the whole register as one pool, ignoring layer, batch and pass: a recall‑oriented matcher proposed every plausible pair within each nation, and each pair above the threshold was adjudicated individually against a written test of what makes two records the same asset‑system. The first merged 412; the second, run after the expansion, adjudicated 6,290 pairs and merged 8, one of them a duplicate caught while rescoring, because the expansion had been given each nation's existing list and told not to repeat it. Nothing was deleted: every merge is itemised in the published ledger, and surviving records carry the absorbed names and ids.
In the twenty first‑pass nations, every national top 100 was then reviewed record by record: name checked against the common or official form, a public reference attached where one exists, coordinates checked and moved where wrong, classification checked against the rules above, score challenged, and evidence confidence set afresh from what the review could actually establish.
What this index deliberately does not do
It does not rank assets by fame, conflate sector size with strategic importance, penalise opaque states for thin public information, double-count one system under several sectors, treat every component of a network as an independent asset, impose equal sector quotas on the national ranking, use tactical vulnerability as a proxy for strategic value, or count planned capacity as current capacity. Every ranking should be reproducible from classification, source evidence, rubric scores and the ranking rule.