Ten measures. Three areas. One rule about evidence.

We only count what anyone else could find too: an original, dated, clearly-sourced public document, published before the cut-off date.

40%

Policy & institutions

Are the rules and institutions in place?

30%

Finance & spending

Can you see climate action in the budget?

30%

Implementation & outcomes

Is the work delivered, and is the result measured?

Code	What it measures	Area
P1	Climate policy / strategy quality	Policy
P2	Climate legal authority	Policy
P3	Functional climate institution	Policy
S4	Executed climate expenditure aligned to plan	Finance
S5	Renewable share of executed energy capex	Finance
S6	Climate-fiscal accountability	Finance
O7	Verified GHG emissions reduction	Outcomes
O8	Unique direct resilience beneficiaries	Outcomes
O9	Priority-group share of beneficiaries	Outcomes
O10	Achievement of eligible due targets	Outcomes

What happens when we find nothing

Where no qualifying primary document exists, the measure is recorded as not shown. It is not scored as a substantive zero and it is not a finding that nothing is happening — it means nothing we could trace has been published. Send us the document and the record changes. That is why the share of the record we found sits next to every score, and why a low score should never be read as “nothing is happening”.

Documents found after 31 December 2025 count in the next round. Each measure is scored 0–3 against a fixed set of rules.

What this does not capture

It rewards publishingWe can only measure what is published, so governments that publish little look weaker than they may be.
Who did the workDonor programmes drive much of the delivery in 33 states. Separating out what the state itself did is ongoing work.
Systems, not resultsThere is not yet enough published evidence of real-world results to tell states apart on that alone.

Any government can send us documents for review at any time. Every change is logged and applied in the next update.