Measurement & Census Method
Every national population figure rests on a definition of who counts as resident. Migrants sit awkwardly against almost every version of that definition.
The residence problem
Censuses count people at a usual place of residence on a reference date. The concept assumes one household, one country, one answer. A substantial share of the world’s population does not fit: seasonal workers, students abroad, labour migrants on multi-year contracts, families maintaining two homes across a border.
Depending on where the line is drawn, the same person may be counted twice or not at all. Because origin and destination countries draw it differently, bilateral migration figures frequently disagree — sometimes substantially.
Why undercounting is systematic
The error is not random. It concentrates among the people hardest to reach:
- Irregular status. Those without documentation have a direct reason to avoid an official enumerator.
- Non-standard housing. Labour accommodation, informal settlements and shared rooms map poorly onto household-based sampling frames.
- Mobility. People who move between sites during the enumeration period can be missed by every pass.
- Language. Instruments administered only in official languages lose respondents who cannot complete them.
Because these overlap, undercounting falls heaviest on the groups whose circumstances most need to be documented.
What improves the count
Administrative data linkage. Border records, work permits, consular registration and remittance flows each capture part of the picture that a household survey misses.
Mirror statistics. Comparing a destination country’s count of migrants from Pakistan with Pakistan’s own estimate of emigrants to that country exposes the size of the discrepancy, even when it cannot immediately resolve it.
Dedicated modules. Adding migration questions to existing household surveys — who has left, when, where to, whether they send money — captures the origin side at modest cost.
Enumerator training. Much of the loss is operational rather than conceptual: people missed because the household was approached once, in one language, at a time when those present could not answer for those absent.
Why it matters
Migration estimates feed directly into population projections. An error in net migration compounds over a projection horizon, shifting the forecast age structure and with it the planning assumptions built on top of it. For a country with persistent net emigration, getting this measurement right is not a technical footnote — it is a determinant of whether the projections are usable.