How it works

What do the labels mean?

Read the evidence label and the robustness verdict together. They describe different aspects of an estimate.

What-if model · Chosen targets, conditional projections and calculated gaps.

First identify the number's role: a chosen target, conditional projection or calculated gap. An E label does not make a target a prediction. The original source labels are preserved, including defaults discussed on the sector pages.

Understand the number's role

Where the number comes from

LabelMeaning
[S] SourcedSourced. A value read from a source.
[E] EstimatedEstimated from sourced inputs. The calculation or interpretation introduces modelling choices.
[A] AssumedAssumed. At least part of the estimate depends on an assumption.

How it behaves in the tests

LabelMeaning
SOLIDThe reported constraint binds across the production alternatives tested. K11 saving has an explicit accounting-identity exemption and only one tested case.
SCENARIOThe constraint depends on a production assumption, or has insufficient alternative tests. Some longer verdicts flag cases that exceed the testing threshold only in another variant.
UNTESTEDThe size or rate depends on an assumption and carries a proposed test. This status can coexist with SOLID.
TESTEDThe exported label indicates that the size was classified as sourced or estimated. The underlying claim is still conditional on the model and its inputs.
Scenario rangeThe lowest and highest blocked-output estimates among the alternatives evaluated for that shortage.

For example, robots are labelled SOLID and A. They constrain output in every tested task split, while their capacity estimates depend on assumptions. Their broad range makes the sensitivity visible.

Other terms used on this website
TermMeaning
GDPThe value of final goods and services produced in a period. The model measures annual world output.
2026 PPP dollarsA common purchasing-power unit used to compare output across economies, valued at the model's 2026 price basis.
Constant pricesA fixed price basis for comparing production across years.
TOne trillion. Displayed output amounts divide the CSV's billions by 1,000.
bnOne billion. The default unit for model output amounts in the v4 CSVs is billion 2026 international dollars.
TargetA chosen what-if scenario: double the model's 2026 world-output estimate and allocate that total among sectors using a selected spending rule. Sector targets are choices, not demand forecasts.
On courseA conditional baseline projection for 2035: today's output plus on-course growth, if the model's assumptions hold. It is not a guaranteed outcome.
On-course growthThe projected increase above today's output under the baseline assumptions. It excludes both calculated tier gaps.
Tier 1Extra output that inputs growing at the model's record pace would make. Closing this gap takes building and scaling up what people already know how to do. The amount adds to the conditional baseline and is capped at each chosen sector target.
Tier 2Extra output beyond the model's historical benchmarks. Reaching the doubling target calls for discoveries that cut the inputs needed per unit of output or break a historical speed limit. The amount is the gap left after baseline growth and tier 1. Today + on-course growth + tier 1 + tier 2 = target.
Worker-equivalentAn estimate of annual work capacity expressed relative to a human worker. Conversions from machines to this unit depend on productivity assumptions.
ElasticityA measure of the proportional response of one quantity to a proportional change in another, used here for income, prices and substitution.
PrecedentA historical rate or ratio selected as a demonstrated comparison. Applying it to the world and to a new period is a modelling choice.

Copy this prompt

Automatic copying is unavailable. Copy the selected text and paste it into your assistant.