Guide · A3

Real vs indicative data

Every country ranking faces the same tension: users want a single sortable number, but jurisdictions publish fragmented, delayed, and politicised information. Liberty Stack Ranking resolves that tension with transparency — not by pretending every cell in the database is audited fact.

This guide separates sourced figures, structured judgements, and missing data; explains what the product will never fabricate; and shows how to read scores without overconfidence.

Three layers of data

Layer 1 — Public anchors

Some fields tie directly to public sources: corporate tax rates referenced against OECD or government tables, CPI (corruption perceptions) in Business mode, IMF or World Bank series where cited in methodology. These anchors ground the dataset but do not automatically populate every score.

Layer 2 — Structured indicative scores

Most Liberté dimension scores (0–10 on tax, money, economy, property, self, space, infra) are editorial judgements encoded consistently across 238 territories. They synthesise law, practice, and comparative experience into one integer per axis. The methodology document states the rubric; the score is indicative, not a statistical measurement.

Business mode adds indicative criteria too — banking friction, reputation, enforcement climate, holding regimes — alongside anchored fields such as headline corporate tax.

Layer 3 — Explicit gaps

When information is insufficient, Liberty Stack prefers unknown or absent over invention. Homeschool status includes an unknown category where territory classification is not yet filled. Real-estate mode may fall back between holding types when a specific cell is empty, with the UI signalling the gap.

What we do not invent: customer testimonials, usage statistics, "thousands of founders chose…" claims, provider rankings, fake pricing quotes, or precision where only bands exist.

Why indicative still helps

Decisions require comparison before precision. Knowing that Panama and Estonia sit in different bands on enforcement and corporate tax orientation is useful even when your exact invoice depends on substance and treaty position.

Indicative scores compress complexity so you know what to research next. They are scaffolding, not verdicts.

Country business sheets pair Ranking data with cost bands (low / moderate / high / very high) and short honest notes — always labelled indicative. Malta may look attractive on tax maps yet carry high operating cost in editorial notes; that is intentional friction, not marketing.

Real vs indicative in each mode

Liberté (seven dimensions)

Dimension scores are rubric-based judgements unless a sub-field (e.g. numeric tax_detail min/max where populated) cites derivable rates. The expat regime toggle swaps in linked regime tax scores — still indicative eligibility and benefit snapshots.

Business (~40 hubs)

Corporate tax and CPI lean on public references; banking, legal, admin, holding, CFC, and enforcement scores are estimated climates for comparison. CFC labels describe how heavy controlled-foreign-company rules feel if you are tax-resident in that jurisdiction — not legal determinations for your structure.

Real estate (~30 freehold markets)

Mix of calculated fields (yield, trend) and stored indicative subscores. Capital-gains treatment varies by holding mode (primary, personal, company); the UI shows which mode is active.

Versioning and change

Scores carry a version stamp (see methodology — currently v10 family). When rubrics shift — for example decoupling self / space / infra from economy — rankings move even if raw country facts did not. That is a feature: the model should improve, not fossilise.

Updates flow from seed-data through validation and reload. If a territory's law changes tomorrow, the database may lag until edited. Scores can lag when law changes until the dataset is updated.

How to read a territory row without fooling yourself

1. Check the mode — Liberté vs Business vs Real estate answer different questions.
2. Open the breakdown — Global rank hides axis lows that veto a move.
3. Read labels — "Indicative," "structured judgement," and categorical homeschool status are precision warnings.
4. Follow sources — Country pages list references where editorial cost notes depend on them.
5. Triangulate — Use Ranking to reorder by your weights, methodology for definitions, and country business data for hub-specific context.

Epistemic hygiene for entrepreneurs

Prefer bands over false precision. A score of 7 is not "70% free." It means "clearly favourable on this rubric."

Separate legality from score. Homeschool may be regulated (legal with friction) while self stays moderate because of conscription or health mandates.

Avoid anecdote wars. Forum stories about one bank branch or one visa officer do not update the database — but they remind you that execution risk sits outside any index.

Next

Pick one decision (residence + company hub). In Ranking, see which axes drive it. For each shortlist — Georgia, Mauritius, USA Wyoming — note which scores are anchored vs estimated.

Confidence ladder

Prefer a short stack — life, business, capital — over omnibus “best country” lists.