Population, land, geography, neighbours, resources, fractionalisation, legal inheritance, starting point. Twenty six facts, unscored, that choose a country's peers and set what should be expected of it.
Benchmarking instrument · not published, available on request
Every country graded against the hand it was dealt
Ask which country is best and the usual answer is a league table that rewards being born rich. This instrument asks a different question. It grades a country twice, on where it stands and on how far that is from what its endowment predicts, and reports the two as a pair. Its founding criterion is self correction: a country where an error has a route to reversal that does not depend on whoever is governing.
The idea underneath it is the same one that runs the Meridian case study, where a hospital is graded per place rather than against its rivals. A country that starts with a coastline, oil, a legal inheritance and friendly neighbours should be expected to do well. The interesting number is not whether it did well. It is how far it landed from where its start should have put it.
So the instrument builds the expectation first, from twenty six facts nobody can take credit for, and only then scores the things a government can move.
26facts of endowment that build the peer group and the expectation, none of them scored
9floors assessed pass, flag or breach, each one capping the composite
102indicators across ten pillars, each with its source and a robustness class
2grades per country: the absolute level and the residual against the hand it was dealt
Chapter 1 · How a grade is built
Four layers
Nine conditions a country either meets or does not: pass, flag or breach. A breach caps the composite, because no amount of growth repairs a floor. Which floors count is a values decision and is stated as one.
Ten pillars, 102 indicators. Every indicator carries its source and a robustness class: hard, survey, perception or state reported, so a reader knows how much to trust each number.
Where it is, where it has been, where it needs to be. Gaps are converted to closure dates, so a shortfall becomes a year rather than an adjective.
Chapter 2 · Two grades, reported as a pair
The central move
Every country is graded twice. The absolute level says where it stands. The residual says how far that is from what its endowment predicts, above or below. A rich country below its own line and a poor country above it are both visible, which a league table cannot show.
The comparison the instrument sets up is never the strongest country against the weakest. It is a country against the peer that started with the same hand and used it better, which is the only comparison a government can act on.
Chapter 3 · Nigeria, as a worked example
The first run
The first country run was Nigeria. On the demonstration loading it stands last of nine on level, misses what its endowment predicts by about half as much as the United States misses its own, breaches seven of the nine floors, and has slipped since 2016 on eight of ten pillars, with state capacity as the binding constraint. On external footprint it scores above Switzerland and the United States. The comparison it sets up is Nigeria against Botswana.
Those findings are reported with their basis. The demonstration scores for the loaded countries are indicative judgements, not sourced data, and a real run rebuilds every indicator from its source. That is the work a request buys.
Chapter 4 · Plainly
What state it is in
The instrument is built and in use: the structure, the floors, the indicators with their sources and classes, the time views, and the loading for nine countries. What is not done is the sourcing of every number for every country, which is a country by country job.
It is not published. If you need one country graded properly from its sources, several compared on the same basis, or the instrument itself with your own floors, ask. The first conversation costs nothing.
It is not published, but it exists.
Name the country or the comparison, and I reply within one working day.
Ask about the Country Benchmark Tool