Method is the product.
In a market this data-constrained, the conclusion is worth nothing without the procedure that produced it. This page is the procedure — source hierarchy, evidence tiers, how a fact is distinguished from an estimate, how confidence is assigned, and what we do when the data simply is not there. It applies to everything Nyra publishes.
Every load-bearing number carries a source and a tier
Not a bibliography at the end — the tier travels with the number, in the sentence where it is used, so a reader can see instantly whether they are looking at a census count or a broker quote.
Conflicts are recorded, not averaged
Where two official sources disagree — and in the Maldives they frequently do — both are carried, the reason for the divergence is stated, and no midpoint is invented. Averaging away a real measurement conflict destroys the most useful information in the data.
Absence of data is a finding
There is no Maldivian property transaction register, no rent index, no wage series and no income survey since 2019. We say so, and we say what the missing data would change, rather than filling the gap with a plausible number.
Estimates declare their vintage
A figure combining a 2019 unit value with a 2026 population is a volume at 2019 prices and is labelled that way. Silent vintage mixing is the most common route to false precision.
Models must beat an honest baseline
No forecast is published unless it beats the seasonal-naive baseline out of sample. Backtest metrics are embedded in the artifact, and uncertainty bands are part of the product rather than a disclaimer under it.
Limitations are published with the work
Every paper ends with what we do not know. That section is written before the conclusions, not after, so it constrains them.
Sources are ranked before research begins, not chosen afterwards to support a conclusion. A lower-ranked source may be used where nothing higher exists — but it is tiered accordingly, and the conclusion's confidence falls with it.
Maldives Bureau of Statistics — Census, HIES, Labour Force Survey, CPI, national accounts, population projections.
Maldives Monetary Authority, Ministry of Finance, MIRA — monetary and payments statistics, debt bulletins, budget documents, revenue data.
Tourism, Housing & Urban Development, Economic Development, Immigration — registers and administrative counts, used with their coverage stated.
IMF Article IV, World Bank Maldives Development Update and WDI, ADB, ILO, UN agencies — including modelled estimates, always labelled as modelled.
Primary law and regulation, quoted and linked rather than characterised.
Audited financial statements, product terms, disclosed project figures.
Peer-reviewed work, weighted by method rather than conclusion.
Developer prices, broker quotes, listings — used as clearly tiered market evidence, never as statistics.
Reporting from identified outlets; expert opinion identified as opinion.
What kind of thing a number is. Attached to the number itself, everywhere it appears.
| Tier | Meaning |
|---|---|
| OFFICIAL | Published by a government body or international financial institution. |
| TRANSACTION | A verified completed deal. Extremely scarce in the Maldives, and treated as such. |
| ASKING | A developer or broker asking price. Not transaction evidence. |
| LISTING | An advertised rent or price. Advertised, not contracted. |
| INDUSTRY-EST | An estimate from an industry participant or trade body. |
| MEDIA | Reporting from a named outlet. |
| COMMENTARY | Expert or market opinion, identified as such. |
| OWN-CALC | A Nyra calculation, model or assumption — method always stated. |
What kind of claim is being made. A figure can be an official fact, an own-calculation estimate or an own-calculation scenario; the two labels answer different questions.
| Class | Meaning |
|---|---|
| Fact | Directly observed and published by a primary source. |
| Estimate | Derived arithmetically from published figures, with the derivation shown. |
| Inference | An interpretation the evidence supports but does not demonstrate. |
| Scenario | A forward path under stated assumptions. Not a forecast unless a model and a backtest exist. |
Every paper carries an overall rating, and important findings carry their own. The rating is about the evidence, not about how strongly we hold the view.
| Rating | Means |
|---|---|
| HIGH | Strong primary data, corroborated across independent sources, with no material definitional dispute. |
| MEDIUM-HIGH | Good primary evidence with a known limitation that does not change the direction of the conclusion. |
| MEDIUM | Sound evidence with meaningful gaps — for example a key input that is several years old. |
| LOW-MEDIUM | Indirect evidence, or a conclusion resting on modelling assumptions we cannot test. |
| LOW | Limited observations or high modelling uncertainty. Published because the question matters, flagged because the answer is weak. |
Research is filed under the market it primarily addresses, and cross-listed into every market it materially informs. A paper on the fiscal position written for the real-estate programme is Economy research; a paper on population and incomes is People research. The relationships are declared in each publication's metadata rather than hard-coded page by page, so a new paper joins the network the moment it is published.
That structure exists because the analytically interesting questions sit between markets: population and real estate give housing demand; population and business give consumer demand; infrastructure and real estate give land value; migration and housing give rental demand. Treating the markets as silos is how most analysis of small economies goes wrong.
Quantitative artifacts — series, forecasts, model outputs — are produced by a versioned pipeline. Each is validated against a schema, carries lineage identifying the run, the code revision and the input hash that produced it, and is immutable once published: a change means a new version, never an edit in place.
Programme-level research dossiers sit behind every paper. Each domain is investigated independently against primary sources, and records its own findings, source conflicts, data gaps and full source list before any paper is written. Every load-bearing figure in a paper is audited back to its dossier entry.
Programme-specific method — research questions, domain decomposition and the analytical models used — is published alongside each programme. The real-estate programme's is here; the domestic-demand framework is here. The publication contract for datasets and the standard any Nyra index has to meet before it carries a number are on Data & indices.
Nyra publishes independent market and economic research. This material is general information and commercial analysis only. It is not investment, legal, tax or accounting advice, is not a recommendation to buy or sell any security or asset, and does not take account of any reader's objectives or circumstances. Nyra is not a licensed investment adviser. Figures are sourced and tiered; estimates and scenarios are labelled as such and may change as new data is published.