Read beyond the noise.

A city’s population depends on where the boundary is drawn

A view of a city spreading into suburban neighbourhoods from a hillside.
AI-generated photographic illustration; not documentary photography.

One article says a city has a million residents; another gives several million. Before assuming one number is wrong, check whether the two sources have drawn the same boundary.

The United Nations’ World’s Cities data booklet explains three useful concepts. A city proper follows an administrative boundary. An urban agglomeration follows the connected built-up area. A metropolitan area takes account of wider economic and social connections, such as commuting. This is a reference for the concepts, not a source of current population rankings.

Three questions, three possible counts

Suppose an imaginary central municipality contains 600,000 residents. Its continuous urban development crosses into neighbouring municipalities, while a wider belt of towns sends commuters to jobs across the region. A count for the central government area, the built-up settlement and the commuting region could all be valid. They would answer different questions.

A report about the municipal library budget may need the first boundary. A study of urban land cover may need the second. A transport discussion may need the third. Choosing the largest number because it sounds more impressive would not improve any of those analyses.

Boundary changes can also affect a time series. If a municipality annexes a neighbouring area, its reported population may jump even without the same number of people moving there during the year. Look for notes about the geography as well as the census date.

Read the footnote before sharing the ranking

  • What geographical unit does the source name?
  • Which year does the figure describe?
  • Is it a census count, an estimate or a projection?
  • Do all entries in the comparison use compatible definitions?

Keep the unit beside the number in your own notes. “Metro area, estimate for year X” is much harder to misread than a bare figure copied under a city name.

This does not make city statistics unusable. It makes them more useful: the boundary becomes part of the information, allowing you to judge whether the count fits the story being told.