The supply shock cascade
Knock out one supplier, whether a factory, an oil field or a mine, and the shortfall doesn’t stay put. Pick a disruption, set how severe it is, and watch the output loss ripple outward, tier by tier, through every customer industry it feeds, traced through the OECD’s input-output tables.
Estimated output loss by industry, by distance from the disrupted supplier
Notes: the centre is the disrupted supplier. Each ring is one more step down the supply chain:
tier 1 buys directly from it, tier 2 buys from tier 1, and so on up to 4 steps; a link counts only if the seller sells at
least 2% of its output to that buyer. Dot area and shade scale with the industry’s estimated output loss, in % of
its output. Estimates come from the Ghosh supply-driven input-output model and assume buyers cannot switch supplier or
draw down stock in the short run, so they are upper bounds. Input-output structure of 2020.
Source: OECD Inter-Country Input-Output Tables (2020, 77 economies, 45 industries); Simdots calculation.
Who feels it, by economy
Estimated output loss, economy-wide average, %
Notes: output-weighted average of the estimated output loss across each economy’s
industries; same model and assumptions as above.
Source: OECD Inter-Country Input-Output Tables (2020); Simdots calculation.
| Economy | Output shortfall (%) |
|---|
Source & method
Model: the Ghosh supply-driven (allocation) input-output model, the forward-linkage mirror of the Leontief price model used on our commodity cost ripple page. A supplier’s output cut is allocated to its buyers by their existing purchase share, and each buyer that can’t make up the shortfall from elsewhere passes its own output cut on to its buyers: dx = dx₀·(I − B)⁻¹, B being the share of each sector’s output sold to every other sector (Miller & Blair, ch. 11).
Tiers: computed by breadth-first search over the same allocation graph, capped at 4 hops; a sector’s tier is the shortest path from the disrupted supplier, following only links where the seller sells at least 2% of its output to the buyer (to keep the diagram to the links that actually matter).
Data: OECD Inter-Country Input-Output (ICIO) Tables, 2020, public edition: 77 economies, 45 sectors. Coefficients computed once and shipped as CSV / JSON.
- Illustrative, worst case. Assumes buyers cannot substitute to another supplier or draw on inventory, fixed technical coefficients, and a short run; real supply chains absorb a good share of a shock this model shows as a full pass-through.
- Disruption scenarios are a handful of recognizable country-sector pairs, not every possible shock; the underlying model covers all 3,465 OECD ICIO country-sectors.
- The map shows economy-wide average output shortfall (output-weighted across a country’s sectors); the bars and ring tables show individual country-sectors.
Licence: this page and its derived data are CC BY 4.0. OECD ICIO used under its public-edition terms.