Loading the data page
Data
The rules that keep official values intact and comparisons honest. For the indicators, releases and sources themselves, see the data catalogue.
Official statistical values come from their named publishers, retrieved at a recorded moment, archived unchanged, and pinned to one publication date so a chart drawn today can be reproduced tomorrow.
We record the publisher, its licence and the credit it asks for before any value is read.
We keep the exact file the source served and the time we retrieved it, so every value can be traced to what the source published.
We record explicit permissions, restrictions and cases where no terms are stated. Each use follows that publication’s recorded conditions; third-party material retains its own restrictions.
Before two series share a chart, we check they measure the same thing.
We store each value as published, together with the date the source published it. A revised release does not overwrite the earlier one: both are kept, so a figure quoted from this site remains checkable after the source revises it.
Transformations such as an index or a growth rate are computed on demand and labelled. The original level is always available in the data table beneath the chart, and in the explorer.
Context metrics are shown at their own cadence and scope. Some are annual totals, shown where a monthly series is not yet approved; others, such as Malaysia's border arrivals, measure a related but different concept from the main tourism series. Either way, the latest approved value is shown as published, and nothing is filled forward.
Partial period: this value covers part of the period, so it is excluded from growth and comparisons.
Spotted a value that looks wrong, or have a question about how something was put together? Email singapore@horwathhtl.com.
Countries count visitors and hotels in different ways. No chart on this site places two series side by side unless they measure the same thing, and where they do not, we say so and show each on its own.
| Result | What it means |
|---|---|
| Directly comparable | The series share definitions, coverage, units and frequency, so levels can be shown together. |
| Comparable after transformation | Levels differ in basis, and the stated transformation makes them comparable. |
| Directionally comparable | Movements can be compared as an index or a growth rate. Absolute levels are not equivalent. |
| Not comparable on one chart | Definitions or periods differ in ways no transformation resolves, so we show separate panels. |
| Comparability unknown | We cannot prove compatibility from the documented metadata, so no combined view is offered. |
For analysts
The labels and rules behind the badges on this site, for anyone checking our work in detail.
A market appears here from the moment we catalogue a source for it, not from the moment we publish a value.
Publication cadences are our estimate from the source's own pattern, not a commitment the source has made. We allow the usual lag after a period closes, then a grace window, before calling a release overdue.
Accessible is not the same as republishable. Each source carries a rights record for every way we might use it, and explicit restrictions or conflicting notices block the affected use. No terms stated is recorded separately from terms we could not inspect.
Some official publishers state no reuse terms at all. We treat that as an absence rather than a prohibition, and publish those values with the publisher credited, a link to the page they were read from, and a commitment to withdraw them if the publisher asks. Where a publisher reserves its rights or requires written consent, we publish nothing from it and say so on its page. Request a correction or withdrawal.
When two series cannot share a chart, the chart names the reason. These are the reasons it can give.
| Reason | Explanation | Code |
|---|---|---|
| Origin concept differs | One series counts by nationality or a source-reported total, the other by country or place of residence. | ORIGIN_NATIONALITY_VS_RESIDENCE |
| Visitor definition differs | The series do not define a visitor the same way, for example overnight tourists against all international visitors. | TOURIST_VS_VISITOR |
| Same-day visitors included in one series only | One source counts same-day visitors and the other does not. | EXCURSIONISTS_INCLUDED |
| Transport coverage differs | The arrival modes counted are not the same. | AIR_ONLY_COVERAGE |
| Border coverage differs | The entry points counted are not the same. | LAND_EXCLUSION |
| One series counts border crossings | Border entries are not the same measure as tourism arrivals. | BORDER_ENTRY_NOT_TOURISM |
| Accommodation universe differs | The hotel statistics cover different sets of properties. | DIFFERENT_ACCOMMODATION_UNIVERSE |
| Method differs or series break | The series come from different methodologies, or a source has flagged a break in the series. | METHODOLOGY_BREAK |
| Collection method differs | An administrative count is compared against a survey or an estimate. | COLLECTION_METHOD_MISMATCH |
| Price basis differs | Current prices are compared against inflation-adjusted values. | NOMINAL_VS_REAL |
| Currencies differ | We do not convert official values between currencies. | CURRENCY_MISMATCH |
| Frequencies differ | Monthly, quarterly and annual series are not placed on one axis. | FREQUENCY_MISMATCH |
| A period is incomplete | One period covers part of its span and cannot be compared as a full period. | PARTIAL_PERIOD |
| Seasonal adjustment differs | One series is seasonally adjusted and the other is not. | SEASONAL_ADJUSTMENT_MISMATCH |
| No shared base period | An index cannot be built without a period both series cover. | MISSING_COMMON_BASE_PERIOD |
| Definition not documented | The source metadata does not establish what is being counted. | UNKNOWN_DEFINITION |
| Units differ | The series are published in different units. | UNIT_MISMATCH |
| Population covered differs | The series cover different populations. | UNIVERSE_MISMATCH |
| No common complete period | No period exists where every selected series has a complete value. | NO_COMMON_COMPLETE_PERIOD |
| Base period value is zero | Growth or an index cannot be computed from a zero base. | ZERO_DENOMINATOR |
| One series contains another | A selected region already includes a selected market within it, so one chart would count the same visitors twice. | HIERARCHY_OVERLAP |
| Dimensions differ | Purpose, mode, visitor type, demographic, accommodation class, expense or subnational cuts do not match. | DIMENSION_MISMATCH |
Where we compute a figure rather than read it, we label it as ours and publish the formula. A derived value is only available where every input permits it.
Room revenue (S$ thousand) × 1,000 ÷ available room nights
Computed by Horwath HTL from official inputs.
Each of these is published with its formula when the indicators it needs exist for enough destinations to make it meaningful. We would rather show nothing than an index nobody can reproduce.
| Index | What it measures | Formula sketch | Blocked on |
|---|---|---|---|
| Tourism Momentum Index | Short-run direction of visitor demand against its own recent trend. | Rolling 3-month arrivals against the rolling 12-month average, indexed. | Monthly arrivals for at least three destinations. |
| Source-Market Concentration | How dependent a destination is on a small number of origin markets. | Herfindahl-Hirschman index over arrivals by origin market. | An approved release carrying an origin dimension. |
| Seasonality Index | How uneven demand is across the calendar year. | Coefficient of variation of monthly arrivals within a rolling year. | Two full years of monthly arrivals per destination. |
| Visitor Yield | Spend generated per visitor. | Visitor receipts divided by visitor arrivals, in the source currency. | Receipts and arrivals at a common frequency for one destination. |
| Tourism Intensity | Visitor volume relative to the resident population. | Arrivals divided by resident population. | A licensed population series. |
| Regional Dispersal Score | How far demand spreads beyond the primary gateway. | Share of arrivals or room nights outside the capital region. | Sub-national tourism statistics. |
| Hotel Demand-Supply Balance | Whether room demand is growing faster than room supply. | Growth in gross lettings less growth in available room nights. | Both hotel series for at least two destinations. |
| Airlift-Demand Gap | Whether seat capacity is running ahead of or behind visitor demand. | Growth in scheduled seats less growth in arrivals. | A licensed capacity source; public data does not carry scheduled seats. |
| Data Quality Score | How complete, current and well-documented a destination's data is. | Weighted score over coverage breadth, freshness, licence clarity and definition documentation. | Computable now from the coverage matrix; held until the weighting is published. |
| Recovery Index | Current demand against the same period before the pandemic. | Value indexed to the same period of 2019. | Published today as the 2019 index view rather than a separate score. |