Method
How each claim is checked, and the rules every page follows.
From claim to band
The claim
Every claim in the speech is in a ledger with its page and a short verbatim quote, machine-checked against the official text (SHA-256
522766ff42a9…).The baseline
A dated source value selected for the scenario, which may be a historical record or forecast rather than the latest observation. Two independent origins are the preferred standard; where that is not met, the figure is provisional and its limitation is shown. Three outlets repeating one ministry figure count as one.
The record
The historical series, with drought years left in.
The required pace
The growth a year needed to go from the baseline to the target by the assumed year. The speech gives no year, so the default is the end of 2031, adjustable on each page.
The best pace on record
The fastest compound growth over the best 5-year window (3-year if no 5-year window exists). Windows that start below 25% of the baseline are left out, so a rebound from a collapse is not used as the comparator. The 25% cut-off is an analyst choice, shown beside its alternative on every page.
The band (proposed)
Required pace divided by the best pace: up to 1x is on pace, 1x to 2x is stretch, above 2x is beyond precedent, and no usable baseline is not assessable. A band states arithmetic. It is not a forecast and never a probability.
on paceproposed 1x or lessstretchproposed 1x to 2xbeyond precedentproposed above 2xnot assessableproposed no usable baseline
Status: the band rule and the 25% rule are proposals, not yet adopted. This page will say when they are adopted.
The rules every page follows
Evidence, not truth
Raw research is evidence, not truth: it is kept untouched, and only checked records reach the pages.
Never averaged
Conflicts between sources are reported, never averaged.
Three tags
Va primary source opened by a research agent
La lead (secondary, or not opened)
Dour arithmetic on cited numbers
A record marked re-opened was opened again by us, with a log entry.
Every claim, three checks
Every claim we make carries a confidence, what would make it wrong, and, where it is a judgment, a competing explanation.
Open maps
Maps, when they appear, use open data only, with credits.
What would need further checking
Working analysis
This public edition presents working analysis. Verification notes beside headline figures identify unresolved definitions, dates, secondary reporting and missing official series. Calculations depend on the reliability of their inputs. A logged source recheck confirms a reading of that source, not independent measurement of the underlying fact.
Provisional choices
The band labels and cut-offs are provisional analytical choices. A historical growth rate is not a physical ceiling, and a stable band is not proof of feasibility. No likelihood of success is estimated. Missing evidence means not established by the sources reviewed, not proof that an activity did not happen.
What checks prove
Automated checks test quotations, provenance, arithmetic and reproducibility. They do not certify the truth of source material. This release has undergone editorial and technical review; an independent legal review is not recorded.
What this edition covers
Covered, and not
The eight target pages, source registry and historical comparisons are the first edition. The cross-cutting economy row is a nominal US-dollar GDP comparison; exchange-rate movements can change it without the same change in real output. A full macroeconomic model, climate-shock simulation and national capital-allocation model are not provided.
Questions, not recommendations
Published delivery questions are provisional questions for investigation, not verified recommendations. Private discussion cost inputs are excluded. Figures retain their original as-of dates; publication does not refresh them.