Sovereign Infrastructure Vulnerability Index
How SIVI measures a country’s structural exposure to infrastructure disruption — and how to read its ranks.
What SIVI measures
SIVI scores every country on how structurally exposed it is to disruption in the systems it depends on to function — where its energy comes from, how reliant it is on maritime shipping lanes, and how concentrated its trading relationships are with any single partner. It is not a measure of infrastructure quality or investment — a country can have excellent infrastructure and still score as highly exposed if that infrastructure depends on a small number of external sources. A higher composite score means higher vulnerability — countries are ranked with the highest score as #1, so #1 is the most exposed country in the index, not the least.
The three pillars
Energy Dependency
33.33%Consumption-weighted average of import dependency across five fuel types — coal, natural gas, petroleum, nuclear, renewables — so a country’s mix of energy sources determines how much weight each fuel’s dependency carries.
Maritime Exposure
33.34%Inverted Liner Shipping Connectivity Index — fewer, less‑frequent shipping connections score as more exposed. Landlocked states receive a structural zero, the reference case in the shared standard’s own Structural Zero Rules — not a gap, and not a deviation.
Single‑Supplier Concentration
33.33%Herfindahl–Hirschman Index of import‑partner concentration — how much of a country’s total imports come from its largest trading partners, versus being spread across many.
Composite score
Each pillar is percentile-ranked against all countries with data for it, then combined by weight — the shared normalization and weighting rules are covered in full on the Methodology Standards page. For SIVI specifically:
Coverage & partial ranks
SIVI requires at least 2 of its 3 pillars to publish a rank — the full Full/Partial/Insufficient system and the injection algorithm behind partial ranks are covered on the Methodology Standards page. In practice for SIVI:
Data quality output
Every country’s published record carries a per-pillar quality summary — coverage, staleness, and the good/aged/stale mix behind it — following the provenance standard above. Full schema details are in the developer documentation, not published here.
Data sources
| Pillar | Primary source | Refresh cadence | Coverage |
|---|---|---|---|
| Energy Dependency | U.S. Energy Information Administration (EIA) | Daily (cached), weekly (live) | ~217 countries |
| Maritime Exposure | World Bank WDI — Liner Shipping Connectivity Index | Weekly | Coastal states + 44 structural‑zero landlocked states |
| Single‑Supplier Concentration | UN Comtrade — total import partner data | Weekly | Countries with an active Comtrade reporter code |
Deviations from BMS‑1.0.0
SIVI is BMS‑1.0.0 conformant with two documented deviations:
Energy pillar — derived composite, not a single series
BMS‑003 assumes one series per indicator. Energy Dependency is instead a consumption-weighted composite of five separate EIA fuel series, combined before percentile ranking. Documented here per BMS‑003 rule 6.
HHI pillar — extended staleness window
BMS‑005 guidance is a 2-year staleness threshold for annual data. Trade-data reporting lags mean many valid reporters would be excluded at 2 years, so this pillar accepts a wider window, always preferring the most recent year available.
Not a deviation: the landlocked-country structural zero on the Maritime pillar is the reference example in BMS‑002’s own Structural Zero Rules, not a departure from it.