Configuration and performance kept apart. Every number carries its source, vintage and quality grade. No overall ranking — because the weights across layers are value judgements.
Both axes are descriptive — neither pole is better. Pick any two of the nine axes. Point size is population; colour is the cluster the data found, not the archetype I assigned.
The same 20 countries positioned by the two directions that carry the most variance across all nine axes at once.
Positive bars convert input into outcome better than the fitted relationship implies.
swipe charts sideways →
Territorial emissions are what a country emits inside its borders. Consumption emissions are what its consumption causes anywhere. The gap measures how much impact has been exported — and it is the difference between a real ecological achievement and an accounting one.
Share of tested social foundations a country falls short on, against the share of tested ecological ceilings it overshoots. Bottom-left is safe and just; almost nobody is there.
Pick two to four countries. Near pairs teach more than global rankings: holding most things constant isolates the one variable that differs.
Level is the mean of the three most recent available years. Change is against the three-year window centred ten years earlier. Grade: A harmonised and recent, B older or patchier. Nothing is interpolated.