Settling in scores 90, family rights 40. State school for a newcomer's child is paid: 1,035 Singapore dollars a month for primary and 2,190 for secondary, and a place is not guaranteed.
The place interval 37-53 shows where the country lands when the data is nudged and the index is recalculated. One place of difference in the table means nothing; a difference of a league means something.
The thin mark on each bar is the average across 97 countries
The strongest category here is Settling in: 90 out of 100, against a 97-state average of 69.5. The weakest is Family rights: 40 against 72.2. Above the world average: 3 out of 7.
All 97 states are calculated for one applicant: 35 years old, income of $2,500 and $3,856 a month before tax, passport not counted, the most accessible route that is actually in use.
Your passport, your income and your family do not change the table. They change which of its lines apply to you. A different income makes entry read differently, a family makes the family category read differently, and your passport makes everything read differently.
97 states, 26 indicators, seven categories, 2,522 scores. Every score has a source and a source class: primary means the texts of laws and government portals; reliable secondary means reports by international organisations, tax reviews by audit firms and specialist legal publications; weak means community posts and single secondary sources. The classes add up to a confidence level for the country, and a country with low confidence gets neither a score nor a place.
Every indicator is scored against a fixed ruler, not against its neighbours. Adding new countries therefore does not move the old ones, and a change in a score means a change in policy and nothing else.
A high place does not mean a good country. It means one thing: the state machine treats newcomers predictably and by the rules. Living there may be expensive, dull or hot; that is not our question.
We make mistakes and we will go on making them: policy changes faster than any database. Our answer is not a pose of infallibility but a visible trail of corrections. Every correction is published: what it was, what it became, the source and the reason for the error. The corrections log opens with the second issue, together with the first block of changes.
