The Ledger · Medium — two to fifteen years

From The Rational Prophet (Book Four of The Laws of Game Theory)

The bet

The bet

Political participation falls across the band without a matching rise in repression. The incumbents did not have to suppress anyone; expected futility did the work.

Win

Average participation across the country set falls by at least 0.05 on the index, while average repression does not worsen by one standard deviation.

Lose

Average repression worsens by one standard deviation or more — then it is suppression, not demoralisation — or average participation does not fall by at least 0.05.

Expiry

Ten years, read annually.

Check it

V-Dem's participation index against its index of civil-society repression. Annual, free.

In protocol form

Word for word from Appendix D of the book, the Prediction Protocol, where this bet is entered as Participation. References to sections (§) are to that appendix.

Universe

every state, as defined in §3.10, carrying a score on both indicators in each of the ten most recent years in the democracy index edition current at publication, produced by the rule in §3.10 and printed with the baseline.

Source

the V-Dem (Varieties of Democracy) dataset of the V-Dem Institute at the University of Gothenburg, annual and free: the participatory component index alongside the civil society repression indicator, both as published in the edition named with the baseline. Survey series are used for illustration in the text only, never as the basis for scoring.

Definitions

demoralisation means falling participation without rising repression — behaviour, not sentiment. The repression indicator is coded so that a higher score means less repression, so a rise in repression is a fall in that score. A matching rise in repression means the average repression score falls by at least one standard deviation over the same period, the standard deviation taken across the states of the set in the baseline year. It is the sample standard deviation: the sum of squared deviations is divided by the number of states less one.

Calculation

the verdict is on the unweighted average across the country set, not on a count of countries, because the bet is about participation across the set rather than in any one state. Participation falls if its average in the latest year of the edition current at the reading is at least 0.05 below its average in the baseline year, on the index's own scale of 0 to 1; a smaller fall is noise, not a result. The baseline year is the latest year in the edition current at publication. A state missing from a later edition is scored at its baseline values, which counts against the bet, never for it.

Baseline

both indices per state for the baseline year, with both averages and the sample standard deviation.

What was traded away

the original version measured sentiment, which is what the mechanism describes, but survey waves arrive years apart and question wording changes.

Baseline at publication

Reading date of record: 10 October 2026

What is recorded

Participation and repression indices, per state, for the baseline year · both averages and the sample standard deviation · the set of states

Value at publication

V-Dem Country-Year dataset, Core, version 16 (March 2026), the edition current at publication; baseline year 2025. Set: the 174 states, as defined in §3.10, carrying both scores in every year 2016–2025, listed below this table. Averages for 2025: participatory component index 0.4700; civil society repression 0.6521 (higher means less repression); sample standard deviation of repression across the set 1.5791 (read 10 October 2026)

Author’s probability 75% that this bet scores HIT. Registered on 10 October 2026.

Sources

• V-Dem dataset
• V-Dem archive of earlier versions

Not investment advice. The bet describes what the book expects to happen; nothing here is a recommendation to buy, sell or hold anything.

Last reviewed 10 October 2026