Main Findings from a Global Survey of Central Banks Jacqueline Irving, Dilip Ratha, and Sanket Mohapatra Development Prospects Group World Bank Washington DC June 12, 2009

Objectives of the survey Part of a larger cross-country data and information gathering exercise by the World Bank’s Migration & Remittances Team A main aim of the central banks survey is to gain a better understanding of: the regulatory environment for remittance flows; how central banks collect data and other information on migrant remittance flows.

Survey questions – some main themes Which institutions collect data? What are the methods? What RSPs are covered? How are these changing? How are RSPs regulated? What factors affect the choice of remittance channel? How does regulation affect costs?

Current status of survey Two main versions focusing on inward and outward flows Sent to 176 countries worldwide (52 in Africa) 105 countries have submitted survey responses: Of which 32 countries are in Africa Final findings analysis paper is a key input to the larger Africa Migration flagship and is undergoing review/clearance for publication as a World Bank PRWP This survey adapted and significantly extended a World Bank questionnaire completed by central banks in 40 countries in 2004.

Focus on data collection Main findings: Focus on data collection

There appears to be a lack of coordination in data collection Within the same institution Across national institutions Between counterpart national institutions, including for major remittance corridors There appears to be a lack of coordination in data collection: Between the various divisions within individual central banks; Between national institutions responsible for data collection and/or supervisory aspects within a given country; And (somewhat less surprisingly), between counterpart national institutions, including for major remittance corridors For example, one high-income country’s central bank reported time series data for annual remittance outflows in 2002-07 for which it sourced the country’s statistical agency and referred us on to that agency for further information on data collection issues. The country’s statistical agency provided different statistics for the same time series, in submitting a separate response for the data collection questions.

Inflows are better monitored than outflows Almost all (96%) remittance-receiving countries collect data, vs. 84% in remittance- sending countries Data collection going on longer for inflows Data collection is done mostly by central banks in receiving countries; whereas in remittance-source countries, national statistical offices seem more active in data collection. Data collection on inflows ongoing longer than for outflows. 72% of the remittance-receiving countries report that collection of migrant remittances data began more than five years ago. In contrast, for remittance-sending countries, the corresponding figure is much lower at 55%.

*Migrant remittance flows data for 2007. There can be large discrepancies in reporting official data* Georgia: $1.3 bn vs $0.7 bn Latvia - $0.5 bn vs $2 bn Ghana - $100 mn vs $1.8 bn UK outflows: $5 bn vs $15bn There were similar discrepencies between the figures (although not as large) for a few other countries that have since been revised upwards. *Migrant remittance flows data for 2007.

Under-recording of remittances Lower priority given to recording remittance outflows by many migrant host countries—particularly where remittances are small relative to GDP Use of cross-border data reporting thresholds Many central banks until recently relied on data reported solely by commercial banks; more nonbank RSPs are beginning to report data Remittances hand carried by migrants and others visiting migrants’ home countries In-kind remittances Another possible reason for underreporting: central banks earlier relied on data reported solely by commercial banks, and are now making efforts to include transfers through money transfer agencies and other providers of remittance services. Some central banks are also starting to use surveys of migrants in the destination countries and migrant household surveys in the origin countries to supplement the data reported by providers of remittance services.

Informal remittances 42% of the remittance-receiving countries said they estimate informal remittance flows But a considerably fewer number of remittance- receiving countries (17%) provided actual informal flows data estimates Only 4% of the remittance-sending countries (Germany and Russia) indicated that they estimate informal remittance flows 42% of the remittance-receiving countries collect data on informal remittances; 70% do so with some regularity. But less than 17% of the central banks provided actual informal flows estimates in the survey

Methods of estimating informal flows

Better statistics and studies are cited as areas needing attention

Better statistics and studies are even more important for Sub-Saharan African countries

Policy recommendations: Data collection Improve coordination in this area, with more systematic data and information exchange, better communication, and more effective division of labor to avoid duplication of efforts. Better coordination in data collection must occur both across different institutions and between different divisions within the same institution. For major remittance corridors, developing some means of regular, more systematic data and information sharing and exchange on migrant remittance flows between counterpart national institutions. Better coordination in data collection must occur both across different institutions and between different divisions within the same institution in a number of countries. For major remittance corridors, developing some means of regular, more systematic data and information sharing and exchange on migrant remittance flows between counterpart national institutions is also important, in order to ensure the formulation of appropriate national policies, particularly in response to sudden and emerging changes in the size, composition of, and channels for transmitting these capital flows.

Policy recommendations: Data collection Improve procedures for remittances data collection and monitoring, including by compiling and monitoring cross- border remittance flows data at higher frequencies and disaggregated by source country where possible. For many remittance-receiving countries, revise data compilation methods to better distinguish remittance inflows from other capital inflows and by remittance category. As new RSP entrants emerge, it becomes increasingly important for countries to give priority to determining an effective means of taking into account these new channels and technologies in data collection and monitoring.

Next steps for this study Dissemination of survey final findings paper to 176 central banks (imminent)—ahead of publication as Working Paper Final findings will provide input into the Migration, Remittances & Development Project flagship report