Migrant remittances and private capital flows - Which is what?

The Uganda Daily Monitor reported recently that according to the World Bank’s latest Global Economic Prospects report, “remittances to developing countries is forecasted [sic] to recover modestly from $454 billion in 2009 to $771 billion by 2012, which still stands below the 2007 $1.2 trillion.” Since we produce the global remittances data and estimates (which incidentally show that remittance flows to developing countries

Photo © Curt Carnemark /  World Bank

reached $316 billion in 2009 and are forecast to grow by 6.2 percent to $335 billion in 2010 (see Migration and Development brief 12), the numbers in the Uganda Monitor made little or no sense to us. Until our colleague Andrew Burns, the lead author of the GEP 2010, pointed out that the Uganda Monitor has likely mistaken overall private capital flows to developing countries as remittances.

The issue of migrant remittances being confused with private capital flows (and even aid flows) is not new. Central banks all around the world have been struggling with this issue for several years now. A global survey of central banks that we conducted during 2008-09 suggests that central banks find it challenging to separate migrant remittances from other small-value transfers such as trade payments, small investments, and even transfers by/to non-governmental organizations and embassies.

This is arguably the case in African countries, where the data cited by government officials is often significantly larger (and sometimes smaller) than the data reported to the IMF balance of payments statistics. Ghana’s central bank reported in February, 2010, that “Private remittances in 2009 amounted to US$1.57 billion compared to US$1.68 billion in 2008” (see Monetary Policy Committee press release). However, the figure for migrant remittances (the sum of workers’ remittances, compensation of employees and migrants’ transfers) reported to the IMF in 2008 was a paltry $128 million, less than a tenth of that reported by the central bank. Similarly, we found last year that remittances reported by Liberia were larger than its GDP; closer investigation revealed that much of it was transfers to aid agencies and non-governmental organizations (NGOs) and not migrant remittances.   

As part of a World Bank-supported initiative involving CIS countries called the Migration and Remittances Peer-Assisted Learning Network (MiRPAL), one of the issues that is being actively discussed is collecting better data on remittance flows (see GDLN on March 26). At a workshop in Moscow on May 31, 2010, the CIS countries together created an Action Plan for improving remittances data (I had the privilege of helping the participating countries to draft the action plan together with a colleague from the IMF). Some elements of this action plan include improving bilateral cooperation between remittance source and recipient countries, creating a web-based interface for sharing data and metadata, and using household surveys and surveys of migrants to supplement remittances data traditionally collected from banks and money transfer companies. While there are several thorny issues in implementation, such initiatives represent a concrete step for countries taking ownership for improving data on a significant external resource flow that has implications for the welfare of many poor people. The issue of improving remittances data is also high on the agenda of the Global Remittances Working Group, with one of the four thematic groups devoted to this important topic.