WORTH99

METHODOLOGY

How we calculate and estimate wealth rankings

WORTH99 starts with published wealth statistics and can also use carefully validated estimates. We tell you when figures are published, calculated between known values, or modeled from other evidence.

Active routes38 calculator references
Result shapePoint, range or bound
Our ruleLabel estimates clearly
Global modelingIn research, not yet live

ONE RESULT CONTRACT

PositionResult carries the meaning.

The calculator derives presentation from the same authoritative result contract used by comparison, visualization and sharing. Reference geography and unit, support bounds, evidence method, confidence, source identity and display precision travel with the result instead of being recreated in the UI.

REFERENCE UNITS

Household is not a universal label.

WORTH99 preserves the population definition of the underlying route. Household, Family, Economic household, adult and individual references are not silently converted into one another. A comparison is only as meaningful as the reference population it actually measures.

UNCERTAINTY

Estimates come with confidence labels.

Published thresholds, values calculated between known cutoffs, and tested models have different levels of certainty. When the available information supports only a range or bound, we show that instead of claiming an exact ranking.

EVIDENCE ORDER

Direct evidence outranks modeled evidence.

Production-cleared direct or locally reconstructed evidence is preferred when available. Lower-confidence but defensible evidence reduces precision and wording strength. Statistical quality, source rights and confidence remain separate decisions.

ESTIMATING RESPONSIBLY

More countries and useful estimates, clearly explained.

WORTH99 can estimate missing wealth thresholds using validated statistical models and compatible country evidence. Estimated figures must be labeled separately from published statistics, with appropriate confidence and ranges. Top-1% and bottom-1% estimates need additional tail validation. The global fallback is still being researched and will not appear in live rankings until the model and source-rights checks are complete.