Costs
Part of Reading richest rankings with a sceptical eye
4 details of richest rankings mistakes people miss
Richest rankings mistakes readers make: treating a position as precision, an absence as evidence, and a movement in the table as something that happened.
Most advice about rich lists tells readers the numbers are unreliable and stops there. That is true and not very useful, because readers keep reading them. The more practical question is what specific inferences people draw from a ranked table that the table cannot support. There are six, and they are all avoidable.
What to take away
- Rank implies an ordering precision the underlying estimates cannot deliver.
- Absence from a list is a boundary decision, never a finding about anyone.
- A change in position usually means somebody else's figure moved.
The six inferences to stop making
Reading rank as precision. Two entries separated by one position may be separated by an amount far smaller than the uncertainty in either. The ordering presents a strict sequence over quantities that would overlap heavily if their intervals were shown. Sorting numbers is always possible; sorting them meaningfully requires that the differences exceed the error.
Six inferences to stop making
- Rank as precision
- Absence as a finding
- Position change as an event
- Agreement as corroboration
- Round number as measured
Reading absence as a finding. People are missing from these lists for many reasons: they fell outside the population the compiler chose, their assets are structured in ways nobody could see, they live somewhere the compiler does not cover, or nobody thought of them. None of those is a statement about their finances, and absence is silent.
Reading a position change as an event. Rank is relative. A person can drop several places without anything about them changing, because other entries moved. In a crowded band, tiny differences reorder many names at once, and none of it corresponds to anything happening.
Reading agreement as corroboration. When several sites carry a similar figure, the usual explanation is that they descend from a common ancestor. Independent confirmation means separate access to the underlying facts, which almost never exists here.
Reading a round number as measured. Published figures cluster on tidy values because they are outputs of rounded assumptions rounded again for presentation. Real balances do not land there, and the roundness is a signal about the process.
Reading the year in the title as a checking date. It usually records when the page was published or edited. Entries without recent coverage typically carry last year's number unchanged.
What the table does support
There is a small amount of real information in a ranked list, and it is worth naming so the exercise is not purely negative.
What the table can support
What you can take
- Listed people are wealthy
- Multiple compilers agree
- Sectors represented heavily
- Visible from company names
- Compiler's population shape
- Read who is included
- Documented event happened
- Tied to filing
Why it holds
- Listed people are wealthy
- Sectors represented heavily
- Compiler's population shape
- Documented event happened
| What you can take from it | Why it holds |
|---|---|
| That the people listed are wealthy | Multiple compilers reached that conclusion from visible activity |
| That certain sectors are represented heavily | Sector concentration is visible from company names alone |
| That the compiler's population choice was a certain shape | Read who is included and who is not |
| That a documented event happened to somebody | Where the entry is tied to a filing or an announced transaction |
Notice that none of those is a number about a person. Everything the format can genuinely support is categorical.
Where the errors come from
Reader carelessness is not the cause. A sorted table with specific figures asserts precision, and readers respond to that presentation as designed.
The format makes a claim its inputs cannot back, so read the format rather than try harder at the numbers; the commercial reasons it persists are on richest rankings.
A better use of ten minutes
Instead of reading a list, read one document. A company filing, retrievable through the SEC's EDGAR full text search, states what a company reported paying named officers. A survey with a published design, such as the Federal Reserve's Survey of Consumer Finances, states how wealth is distributed across households and admits its own limits.
One tells you something true about a person, and the other something true about a population. Together they cover the honest ground, and neither ranks anybody.
How the same reasoning applies across every field this site covers is on estimate methodology. What the filings genuinely support is on executive net worth, and why lifetime totals are no more checkable is on career earnings.
Common questions
Is there any list worth reading?
Lists of documented, dated facts are worth reading: disclosed pay at named companies, announced transactions with prices, published prize structures. Those are catalogs rather than rankings, and they do not order people by wealth.
What should I do if I find myself or someone I know listed?
Compilers generally accept corrections and some publish a route for it. Whether to engage is a personal decision with real consequences either way, and it is not one a page like this should make for anybody.
Do rankings ever get corrected?
Numbers change between editions constantly. Marked corrections, saying what was wrong and why, are rare, which is what makes it impossible to distinguish a fix from a drift.
