| Other notes | Basis and definition: The data on remittance flows are based on officially reported data in the IMF Balance of Payments. Each country reports these data according to the IMF Balance of Payments Manual 6. Each country reports these data according to the IMF Balance of Payments Manual 6. Remittances are defined as the sum of two components: (a) “personal transfers” recorded under the “secondary income” category of the current account; (b) “compensation of employees,” which includes wages, salaries, and other benefits of border, seasonal, and any other nonresident workers (such as local staff of embassies) and which are recorded under the “primary income” category of the current account.
Remittances are known to be underestimated significantly in the balance of payments statistics of the IMF. There are a couple of reasons that officially reported data on remittance might not correctly capture the full extent of money transfer activities. First, the above definition does not include transfers through informal channels—such as hand-carries by friends or family, or in-kind remittances of jewelry, clothes, and other consumer goods, or through hawala. These are believed to be significant portions of total remittances in many countries. If and when they are recorded, it is not clear to what extent they reflect actual transfers rather than imports. Second, in many countries, nonresident deposits, although classified under the capital account, may in part reflect workers’ remittances. For example, the nonresidential rupee deposits in India are most likely remittances disguised as deposits—upon maturity, they do not return to the nonresident depositor because the rupee is not convertible into hard currency.
Caveats: There are several caveats relating to some adjustments made to the remittance data presented here. The caveats attached to our remittance data are:
(a) For some countries, data on remittance has been missing over several years. Missing value can significantly affect the result of data analysis of our time series data; thus, we use the Last Observation Carried Forward (LOCF) and Next Observation Carried Backward (NOCB) methods. These methods replace missing values either with the immediately preceding observed value or the subsequent observed value. These are potentially useful for time series data but may introduce bias if the data isn’t stationary.
(b) For several countries, including China, Portugal, Spain, and Vietnam, the ratio of personal transfers to the current transfer total under the secondary income has been used in the past to estimate personal transfer data since disaggregated current transfer data is unavailable. However, only what’s presented in the IMF BOP will be used (along with historical data).
(c) For Israel and Zimbabwe, the IMF’s BOP data is complemented with the central bank data. For Israel, personal transfer data from the central bank are being used due to their absence in the IMF’s BOP presentation. For Zimbabwe, remittance data have not been reported in the IMF’s BOP since 2020; thus, central bank data are being used from 2021 onwards.
(d) For the current year estimates of remittance flows, high-frequency data from central banks are being used to fill gaps in data. A few dozen central banks report monthly remittance data, mostly consistent with BOP-reported data. The list includes Bangladesh, Bolivia, Brazil, Colombia, Dominican Rep., El Salvador, Georgia, Guatemala, Jamaica, Kenya, Kyrgyz Republic, Mexico, Morocco, Nepal, Nicaragua, Pakistan, the Philippines, Serbia, Sri Lanka, and Ukraine. |