How to read a migration or travel map: year, measure, unit, and why missing is not zero

A colour-coded map is quick to read, and that can hide detail. A dark patch suggests "a lot", a pale patch "a little", and a grey patch "nothing". In migration data the last reading can be wrong. This guide gives six questions to ask of any colour-coded map, on this site or elsewhere.

1. What is being shaded?

The same colour scale can show very different things: tourist trips, people living in a country who were born elsewhere, or people who settled during a year. These are separate measures with separate definitions.

The UN Population Division, for example, measures international migrants as a stock: the estimated number of people living outside their country of birth at mid-year (1 July) of specific years. It states that its dataset does not measure flows. Eurostat's immigration series is a flow, and only counts people who establish usual residence for at least 12 months, or intend to. A map that mixes the two without saying so does not allow a fair comparison.

2. Which year, and how old is the data?

Many migration figures come from censuses, and the UN documentation says its inputs were mostly from the 2005 to 2014 census round, so countries' reference years differ. Between data points, values are estimated by interpolation, extrapolation and assumed growth rates. A smooth colour gradient across years can therefore hide estimation rather than observation.

Look for the reference year on the map itself. If there is none, treat the map as illustrative.

3. What is the unit?

Check whether the shade shows a count, a share of the population, or a rate. Counts favour large countries. Shares favour small ones. Neither is wrong, but they tell different stories. If a map does not state its unit and denominator, ask before you compare two countries.

4. Who defined "migrant"?

Some countries record place of birth, others record citizenship. The UN documentation says citizenship was needed as a proxy for 46 countries or areas, nearly 20 percent of the total. Under a citizenship test, naturalised people disappear from the count and, under some citizenship rules, people born in the country can be counted as migrants.

Even within Europe, Eurostat lists countries that apply different time rules to "immigration", such as six months instead of 12 in Norway. So two neighbouring shades may not be measuring the same thing.

5. Is grey zero, or unknown?

A blank, grey or missing cell can mean any of these:

• the country did not report;
• the country reported but the series has a break;
• the value was imputed from a model country or region;
• the true value is zero.

Only the last is a zero. The UN documentation says six countries or areas were imputed from model countries or regions, and that data coverage is thinnest in Central and Southern Asia. Eurostat flags series breaks in its database, including a break around 2008 and 2009 tied to a change to the 12-month definition. A well-made map separates "no data" from "low" with a different pattern or label. If a map does not, grey should not be read as zero.

6. Does the map claim more than it can know?

Maps of origin and destination depend on the destination country's records. The UN documentation warns that small origin countries are likely to be undercounted. Eurostat describes mirror discrepancies between what two countries report about the same movement. And census data tends to miss groups such as refugees in camps. A map shows what was recorded, not everything that happened.

How this applies to LetsFlyAbroad

When we show a map or heat layer, we will state the measure, unit, year and source beside it, show missing data differently from low data, and link to the page that defines the measure. We will not show a map where we cannot meet that standard, and we will not use a map to suggest that a route is easy, popular or open to you. Route eligibility lives on route pages and official sources.