The Arrive Index

How the index is built

Short version: 19 indicators, grouped into 6 cohorts, each with a stated weight. Every indicator reads a named field in our own datasets — the programme files, the tax profiles and the country guides — and nothing is scored by hand.

The cohorts and their weights

The indicators inside each cohort

Weights are relative within their own cohort, not out of 100. Changing the persona changes them. A priority toggle doubles the indicators it names and lifts their whole cohort by 60% — without that second step a priority naming every indicator in a cohort would cancel itself out — and the cohort weights are then renormalised to 100.

How the numbers are made comparable

• Continuous figures — rates, rents, months, visa-free counts — are scaled to 0–100 across the 120 countries, clipped at the 5th and 95th percentile first. Clipping stops one extreme rent from flattening the whole column.
• Yes/no facts — a wealth tax, an exit tax, a path to citizenship — score 100 or 0. There is no partial credit for a tax that exists.
• Guide ratings — safety, healthcare, schools, infrastructure — are our own editorial 1–10 assessments, multiplied by ten. They are judgements, not measurements, and they are the softest input on this page.
• Missing values score at the midpoint, not at zero. An absent figure is an absence of evidence, not a bad result. 3 countries currently carry an imputed value.

The gates

Filters are applied before scoring, not inside it, because they are not matters of degree. A country passes a gate if at least one of its live programmes does.

• Capital — the cheapest live route, converted to USD at build time. A route that asks for no capital satisfies every budget. A route with a capital requirement we could not price is treated as failing a cap rather than passing it.
• Time on the ground — classified from the programme's published presence rule into none, a few days, or substantial. It is a classification of free text, not a published number, and where the rule is silent we say so rather than guess.
• A passport, and how soon — the page opens filtered to countries where a genuine route reaches citizenship inside five years, because that is the point at which a second passport stops being a theory. The number of years is the ordinary qualifying period as published, taken as first stated rather than as the shortest: Spain reads ten years for most nationalities and two for Ibero-Americans, and the exception should not be quoted to everybody. Where a country has several routes it keeps the fastest of its own, which is why descent and investment routes show as immediate. A route counts only if it leads to a passport a foreigner can actually obtain. A programme that says none by right, nomination only, effectively none or realistically this leads nowhere does not pass, however long the sentence. Nor does a qualifying period longer than fifteen years: a twenty-five-year clock at a head of state's discretion is a formality, not a plan. Discretion on its own does not disqualify — Singapore and Japan are discretionary too, and they naturalise people every year. This is why the Gulf states rank well on this page and disappear the moment you filter for a passport.
• Keeping your citizenship — read from the programme's dual-citizenship field.

What we deliberately left out

• Whether a country reports your accounts. We hold that flag and we will not score it. Choosing a residence because it fails to send a file is not planning, and rewarding it here would make the index an instrument for something we do not do.
• Passport rankings as the headline. Mobility is in, at six percent, because it genuinely matters at a border and almost nowhere else. An index that leads with visa-free counts is ranking the thing that is easiest to count, not the thing that decides a move.
• Happiness, climate and lifestyle scores. Real for some readers, unscoreable across 120 countries without inventing precision. Climate in particular is a regional fact, not a national one.
• Our own fees, and any weighting toward what we sell. No cohort rewards a country for having an expensive programme.

What this index is not

It is not advice, and it is not a shortlist you should act on unedited. Three limits worth knowing. The daily-life cohort rests on editorial ratings rather than official statistics. Everything is scored at country level, which is the wrong level for cost, safety and schools — those are decided by district, as we argue at length in the piece this index came out of. And the tax cohort scores the system as published, not the position you would actually be in, which depends on where your income arises and what you are leaving behind.

Universe: 120 countries — every jurisdiction where we hold a programme file, a tax profile and a country guide. Programme and tax data carry their own verification dates on their pages. Rebuilt with the site; the ranking above moves when the underlying data does.