How much of what migrants send home does anybody actually observe?
Measuring Remittances
Giuseppe De Arcangelis et al. (2022) · Journal of Development Economics
Article ↗ · Replication data ↗ (openICPSR)
The archive requires an account we do not hold, so the files were not opened. What follows rests on the published abstract and on this platform's own observations.
What the authors found
De Arcangelis, Fertig, Liang, Srouji and Yang tackle the measurement problem directly, comparing survey responses against administrative records and a transaction-logging app for the same individuals. The concern is not academic: remittance statistics underpin development policy, and if surveys and official channels miss a substantial share of what moves, the aggregate everyone cites is wrong in a direction nobody has quantified.
What our data adds
This database carries both sides of the recorded flow, and the two sides do not balance in a way worth noticing. The United States pays out 107.1 billion dollars a year, Saudi Arabia 58.7, Switzerland 42.7, Germany 26.7. On the receiving side India alone takes 150.7 billion. France appears in the top ten of receivers at 40.4 billion and also among the largest payers at 20.8 billion, which is a reminder that these lines capture transfers in both directions for any country with both an emigrant and an immigrant population. Luxembourg pays out 20.0 billion — for an economy of its size, a figure that says more about cross-border commuting than about migration.
The other side of the ledger: where the money is sent from.
| Figure | Value | Indicator |
|---|---|---|
| United States, remittances paid | 107.1bn USD | BM.TRF.PWKR.CD.DT |
| Saudi Arabia, paid | 58.7bn USD | BM.TRF.PWKR.CD.DT |
| India, received | 150.7bn USD | BX.TRF.PWKR.CD.DT |
| Luxembourg, paid | 20.0bn USD | BM.TRF.PWKR.CD.DT |
Every figure here is computed from observations in this database and can be charted, downloaded or queried.
What this cannot show
World totals of remittances received and paid do not reconcile, and nobody expects them to: countries compile them differently, informal channels appear on neither side, and some flows are recorded by the receiver but not the sender. The paper exists because the underlying measurement is unreliable, so quoting these figures to four significant digits would miss its point. Treat them as the recorded portion of an unknown whole.
The data behind it
Personal remittances, paid (current US$)
BM.TRF.PWKR.CD.DT
WBPersonal remittances, received (current US$)
BX.TRF.PWKR.CD.DT