Migration figures are easier to misunderstand than to compare. This guide explains the difference between migrant flows, migrant stocks, asylum applications, refugees, and internally displaced people, then provides a practical workflow for checking definitions, time periods, country coverage, rates, and revisions before drawing conclusions about migration trends by country.
Overview
Migration data does not describe one single phenomenon. It can measure people crossing a border during a period, people living outside their country of birth at a particular date, applications for protection, people recognized as refugees, or people displaced within their own country. These measures overlap in subject matter but are not interchangeable.
The most important distinction is between a flow and a stock. A flow records movement during a defined period, such as a calendar year. A stock records a population at a particular point in time. A country can experience a large inflow without having a large migrant stock, while a country with a large migrant stock may have relatively modest new arrivals in a given year.
Protection-related measures require additional care. An asylum application is a request for international protection; it is not the same as a refugee determination and does not necessarily represent a unique person if the dataset counts applications rather than applicants. Refugees are people covered by a legal or administrative protection category, while internally displaced people have been forced to leave their homes but remain within their country's borders.
For editors, researchers, and publishers, the goal is not simply to find the largest number. It is to identify what each number counts, when it was measured, which countries are included, and whether the result is suitable for the comparison being made.
What to track
Migration data glossary
| Measure | What it generally describes | Common measurement problem |
|---|---|---|
| Migrant flow | People who move into, out of, or between places during a specified period. | Definitions of a move vary. Some systems require a minimum duration of residence, while others rely on registrations, permits, border records, or estimates. |
| Migrant stock | People living in a country or area who meet the dataset's definition of an international migrant at a particular date. | Stock can be based on country of birth, nationality, or another definition. It reflects past movements, deaths, departures, and changes in classification. |
| Asylum application | A formal request for international protection submitted to authorities. | Applications may be counted by application, applicant, case, or registration event. A filing is not the same as a final decision or a refugee status. |
| Refugee | A person included in the relevant legal, administrative, or statistical refugee category. | Coverage can differ by source and may include or exclude people in certain temporary or group-based protection arrangements. |
| Internally displaced person | A person forced or obliged to leave their home who remains within the borders of their country. | Estimates may depend on surveys, registration systems, event monitoring, or modeled assessments. A person may be displaced more than once. |
| Return movement | People moving back to a previous country or place of residence. | A return may be temporary or permanent, assisted or spontaneous, and may not indicate that displacement or migration has ended. |
Variables that make comparisons meaningful
Track the measure's unit first. Is it people, applications, movements, households, cases, or events? Then record the geography: origin, destination, residence, nationality, place of registration, or location of displacement. These labels can produce very different results.
Next, record the time period. A year-end stock should not be compared directly with a full-year flow. Monthly asylum applications should not be presented as an annual total unless the aggregation method is clear. Where population sizes differ substantially, calculate a rate per population in addition to the absolute number. A rate can help compare scale, but it does not replace the underlying count and may introduce its own denominator limitations.
Also track whether the figure is a total, an estimate, a provisional result, or a revised series. International datasets can be updated when new administrative records arrive, historical estimates are recalculated, or definitions and country coverage change.
A repeatable comparison workflow
- State the question precisely. Decide whether you are asking about new movement, the resident migrant population, protection claims, recognized refugees, or displacement inside a country.
- Read the definition. Look for the source's definition of migrant, refugee, applicant, displaced person, flow, and stock. Do not rely on a chart title alone.
- Check country coverage. Confirm that the same countries, territories, and reporting periods are included. A regional total may not equal the sum of the countries shown in a selected chart.
- Separate origin and destination. State whether the comparison concerns where people came from, where they live now, where they applied, or where displacement occurred.
- Align the time periods. Compare like with like: year with year, month with month, and point-in-time stock with point-in-time stock.
- Calculate rates carefully. If using a per-capita measure, document the denominator, reference date, and whether the numerator and denominator cover the same geography.
- Check revisions and breaks in series. Note changes in methodology, reporting practice, boundary coverage, or classification before describing a trend.
- Preserve the original figure. Save the download date, table name, definition, and version or revision note so another reader can reproduce the comparison.
Cadence and checkpoints
Migration reporting follows different publication schedules. Administrative measures such as applications or registrations may appear monthly or quarterly, while population stocks and global displacement estimates may be released less frequently. A responsible tracker should therefore use a calendar rather than assume that every indicator updates at the same time.
For a monthly or quarterly monitoring routine, begin with a publication check: identify newly released data, delayed reporting, provisional values, and revisions to earlier periods. Follow with a definition check: confirm that the publisher has not changed the measure, geographic scope, or counting unit.
At each update, compare the newest period with the immediately preceding comparable period and, where available, the same period in the previous year. Avoid turning one unusual month into a long-term trend. Record missing values rather than silently treating them as zero. If a country changes its reporting system, mark the break in the chart and explain that values before and after the break may not be directly comparable.
Every quarter, review the full metadata: country coverage, source notes, estimation methods, revisions, and publication status. Once a year, rebuild the glossary and chart labels from the current methodology documentation. This is especially important for evergreen pages, because definitions can change even when the article's central explanation remains valid.
How to interpret changes
A rise in a flow may indicate more arrivals, better registration, changes in documentation, or a combination of factors. A fall may reflect fewer movements, limited access to registration, delayed reporting, or a shift in routes. The number alone does not identify the cause.
A rising migrant stock usually reflects accumulated demographic change rather than only recent arrivals. It can increase through new migration, longer residence, births among migrant households, improved recording, or changes in how people are classified. It can decrease through departures, deaths, naturalization, or revised estimates. Avoid describing a stock as the number of people who arrived during the latest period.
Asylum applications should be read alongside decisions, pending cases, and the relevant time period when those data are available. Applications indicate demand for a protection process, not its outcome. Comparing applications across countries also requires caution because access to procedures, registration capacity, processing backlogs, and legal systems differ.
Refugee and internal displacement figures answer different geographic questions. A refugee has crossed an international border under the relevant classification; an internally displaced person has not crossed an international border for that displacement. A person can move between categories over time, and one crisis can create both cross-border and internal displacement. A chart that combines them without clear labels can overstate or obscure the pattern.
When writing an explanation, use language that matches the evidence. “Recorded applications increased” is more precise than “more people arrived” when the dataset measures applications. “The reported stock was higher” is safer than “migration surged” when the series is affected by revisions or classification changes.
When to revisit
Revisit this article and any linked charts on a monthly or quarterly schedule when the tracked dataset publishes new observations. Update it sooner when a source issues a major revision, changes its definitions, expands or reduces country coverage, introduces a break in the time series, or replaces provisional figures with final results.
For editors, the minimum update checklist is practical:
- Confirm the latest publication date and reporting period.
- Check whether historical values were revised.
- Review definitions of flows, stocks, applicants, refugees, and internally displaced people.
- Verify origin, destination, residence, nationality, and displacement-location labels.
- Recalculate per-capita rates using a documented denominator.
- Mark missing, provisional, estimated, and non-comparable values.
- Refresh chart notes, download links, and the data dictionary.
Readers can use the same process whenever new migration figures appear in global news. Before sharing a ranking or trend, ask three questions: What exactly is being counted? For what period and geography? Has the series been revised or defined differently? Those checks are usually enough to distinguish a meaningful change in migration patterns from a change in measurement.
For further country-level work, pair this framework with a guide to comparing migration and refugee data across countries and a structured country risk dossier. The central rule remains constant: compare like with like, show the definition beside the number, and preserve the data context for the next update.