Luxembourg Group 1 Workers' remittances, current private transfers and compensation of employees in the German Balance of Payments Statistics

2 Background ❙ The main source for statistics on international remittances is the Balance of Payments (BOP) ❙ In this regard, BOP compilers have been faced with an increasing interest in remittance data in the last decade ❙ This increasing interest, however, puts some pressure on compilers as data on remittances been judged by users as less reliable compared to other BOP items. ❙ Especially the lack of comparability (usually remittance credits exceed debits) indicate differences in the coverage and compilation What are the reasons for these differences?

3 Background ❙ Generally, the multitude of channels used for sending remittances make the collection of data difficult. ❙ The channels vary from country to country, from remitter to remitter, from the financial system in the sending and receiving country, the convenience and cost. ❙ Furthermore, the small size of the individual transactions makes it very difficult to detect remittances in the frame of the general BOP collection system. ❙ It should be also mentioned, that a certain “lack of interest” in developed countries in this sub-item of the BOP, which is of minor importance for them compared to other items, may contribute to some of the inconsistencies

Luxembourg Group 4 Outline 1. Current data sources and data published for the components of the future items "personal transfers" and "personal remittances" a) Workers' remittances b) Current private transfers c) Compensation of employees (net) 2. Weaknesses of the current system 3.Measures to improve the current statistic 4. Compiling and presenting remittance data

Luxembourg Group 5 1 a) Workers' remittances General remarks The collection of data on workers' remittances is embedded in the general reporting requirements. These requirements stipulate that all payments above the relevant reporting threshold (at present 12,500 EUR) must be reported to the Bundesbank. In Germany, it was assumed throughout the years and regardless of the respective amount of the threshold (DM 500, DM 1,000, DM 2,000 etc.), that remittances of migrant workers to their home countries exceeded the threshold only in rare cases. Accordingly, a method was implemented to estimate these flows for the German Balance of Payments.

Luxembourg Group 6 1 a) Workers' remittances Sources  Monthly collective reports by selected banks. Some of them are branches of foreign banks (primarily Turkish but also from other countries like the Philippines). The reports comprise cash deposits with a beneficiary abroad  In addition, some banks report monthly collective reports about cash deposits or transfers of foreigners into accounts held by foreign banks  Data from the Federal Employment Office about the number of employed and unemployed foreigners subject to social insurance contributions (corrected by cross-border and seasonal workers). To assure that the mentioned bank reports contain also information about cross border payments below the threshold, special agreements between the banks and the Bundesbank were made.

Luxembourg Group 7 1 a) Workers' remittances Calculation method 1.To differentiate between workers' remittances and payments for other reasons, percentage rates for each receiving country are applied to the amounts reported in the case of payments to accounts abroad. 2.Remittance flows per capita are calculated by dividing the reported values by the number of non-residents reported by the labour agency. 3.It is assumed, that every (registered) guest worker in Germany remits a minimum amount of EUR 1,200 and a maximum amount of EUR 2,000 to his home country each year. 4.The average amount per capita is calculated on the basis of bank reports and compared with the minimum/maximum amount. If the reported amount lies in between the min/max amount, the reported value per capita is used. If the reported amount is below/above, the min/max amount is used. 5.Multiplied with the number of registered persons per nationality yields to the total amount per country.

Luxembourg Group 8 1 b) Current private transfers (household to household) General remarks Current private transfers are collected via the regular collection system, i.e. only payments exceeding the exemption threshold of EUR 12,500 are reported. For transactions below the threshold an estimation method is applied, however this method was not specifically developed for private transfers. Depart from this “estimation”, no other corrections/estimations are made.

Luxembourg Group 9 1 c) Compensation of employees General remarks Similar to workers remittances, salaries paid are normally below the respective threshold and therefore not captured by the reporting system. Even if a single transaction exceeds the threshold, reporters are often not aware about their obligation to report because, from their point of view, the payments often take place between two resident accounts (for instance, wages of German employees paid by foreign embassies). Hence, the Bundesbank uses indirect sources to estimate compensation of employees gross and net.

Luxembourg Group 10 1 c) Compensation of employees Sources  Statistics of the Federal Employment Office (number of cross border/ seasonal workers and the respective country of the employee)  Statistics of the Federal Statistical Office (quarterly calculation of average gross income (credits/debits for neighboring countries)  Statistics of the Federal Ministry of Finance (compensation of German employees working for foreign military forces stationed in Germany)  Statistics of the Federal Foreign Office (compensation of foreign employees working in German embassies abroad)  Annual reports of International Organizations (compensation of German employees)  Partner country data for receipts of German workers employed in CH, LUX, NL, FR

Luxembourg Group 11 1 c) Compensation of employees Calculation method 1.FSO calculates an average income, separately for cross border and seasonal workers. The average wage is multiplied with the number of employees (foreigners in Germany, Germans abroad). To this amount the contribution of the employer to social insurance is added. For German receipts from CH, LUX, NL and FR we use partner country information. 2.To come to the net income, the total contribution to social insurance and taxes on income is deducted from the gross value. Finally, a correction for travel is made by the Bundesbank. 3.Regarding income received from foreign military forces and international organizations, no additional corrections are made. In the case of income paid/received by embassies in Germany/abroad, information from the national accounts are used.

Luxembourg Group 12 2 Main weaknesses Workers remittances i.Money flows through informal channels and by illegal workers from countries which are not included in the database of the Federal Employment Office are not reflected adequately. ii.Assumed minimum/maximum amount has not been proofed iii.Estimation method does not take into account the propensity to remit. iv.Current data from the banks do not allow to distinguish in all cases for what purpose the actual payment is made, e.g. for basic needs of relatives or for other purposes like deposits in own accounts. v.No information about workers remittances credits

Luxembourg Group 13 2 Main weaknesses Current private transfers Data source does not assure an adequate coverage of all private transfers. The breakdown by country is distorted as only the information of reported payments above the threshold is used. Compensation of employees Germans who work abroad for more than one year are treated as residents and their salaries are included undistinguishable in compensation of employees.

Luxembourg Group 14 3 Measures to improve the current statistics Remittances: Use of data from money transfer companies (Western Union etc.). Currently, the Bundesbank is checking the usefulness of aggregated information provided to us by the Federal Financial Supervisory Authority. Results: Information from this source could be another piece in the puzzle, for instance regarding the country breakdown, but is not the final solution. However, it is planned to use this source as a starting point to estimate workers’ remittances credits.

Luxembourg Group 15 Measures to improve the current statistics With view to future definition of “private transfers” It will be necessary to collect more detailed sector information to capture household to household transfers. Problems: - The sector in the receiving country is not always exactly known by private declarants (how would a private person sectorise a payment via a non profit organisation?) → distortion of the results - Second, political pressure to reduce the statistical burden for private persons. This could result in a complete exemption of private person from all or most of their reporting obligations → use of secondary information may result in less quality and comparability

16 Collection approaches to obtain data (a) Transaction Reporting ❙ Remittances transferred using international networks, typically from the banking system. ❙ Data can be collected from banks using international transaction reporting systems (ITRS) based on the information given by the sender in the payment order. BANK A ACCOUNT COUNTRY A (sender)COUNTRY B (receiver) BANK B ACCOUNT sender receiver ACCOUNT transfer via international network e.g. SWIFT cash DATA COLLECTION A B C A B C

17 Collection approaches to obtain data (a) Transaction Reporting Strengths: ❙ Timely information as data are registered by ITRS at the moment when a transaction is settled; ❙ Low cost as data on remittances are essentially a by-product of the cross border settlement. Weaknesses: ❙ Does not cover international payments outside the banking system or transfers in kind. ❙ Depending on changing political environment (e.g. discussion in the EU on raising the reporting threshold to EUR to reduce the reporting burden for banks in the frame of the Single European Payments Area (SEPA)).

18 Collection approaches to obtain data (b) Direct Reporting ❙ Remittances sent by using Money Transfer Operators (MTO) ❙ MTOs tend to aggregate and net settlement payments. As such, direct reporting would seem to be an appropriate method for collecting detailed information about individual transactions. MTO ACCOUNT of MTO COUNTRY A (sender) COUNTRY B (receiver) MTO ACCOUNT of MTO sender receiver AGENT of MTO Centra l Cleare r of MTO Information: sender, receiver, amount, etc. cash A B C A B C A B C A B C DATA COLLECTION

19 Collection approaches to obtain data (b) Direct Reporting Strengths: ❙ Low costs, because the number of reporters is normally not large. ❙ Reliable as all relevant information is available and full coverage of remittances routed through the MTOs. Weaknesses: ❙ No information about remittances sent through other channels or transfers in kind ❙ MTOs cannot assure that transactions other than remittances are excluded.

20 Collection approaches to obtain data (c) Household surveys ❙ Remittances sent outside the formal sector ❙ Especially if remittances sent via informal channels, the appropriate way to collect remittance data is asking senders and receivers directly. COUNTRY A (sender)COUNTRY B (receiver) Information: sender, receiver, amount etc. BUS COMPANY Physical transport of the money receiver sender CASH A B C DATA COLLECTION busdriver A B C

21 Collection approaches to obtain data (c) Household surveys Strengths: ❙ Direct control over information collected. ❙ Possible to collect data sent through formal and informal channels. Weaknesses: ❙ Costly and methodologically demanding. ❙ Under-reporting because information requested are considered sensitive by the respondents.

22 Collection approaches to obtain data (d) Data Models ❙ Direct reporting (asking banks, MTOs, senders/receivers) is difficult and impractical for a number of reasons. ❙ If direct measurement is no option, models based on secondary data could be used to estimate remittance. BoP COMPILER BANK SENDER MTO RECEIVER Direct data collection e.g. Administrativ Data: - Data about foreign population - Income etc. Secondary data

23 Collection approaches to obtain data (d) Data Models Strengths: ❙ Estimations are definitely able to cover formal and informal channels. ❙ Low costs as secondary data (eg administrative data) is available at no extra cost. Weaknesses: ❙ Input data can be weak (e.g. if migrant population is partly estimated). ❙ Assumptions about variables which influence the propensity to remit may be problematic (problem to test or verify them in practice) and fixed over time.

24 Further guidance Given the multitude of channels and possible approaches that are available for collecting/estimating remittance flows, the compilation guide of the Luxembourg Group further contains a practical four-step approach to develop a country-specific data development plan: Step 1: Understanding the current situation. Step 2: Evaluation of current data collection system. Step 3: Prioritization. Step 4: Improvement and Expansion of data by channel.

25 Further guidance Besides providing definitions, compilation guidance and proposals how to improve data the guide also reviews good practices in disseminating remittance data. ❙ As a general rule, the DQAF of the IMF gives good guidance on diverse aspects of dissemination practices ❙ Further recommended is the publication of bilateral data. However, it is recognized that the collection of bilateral data is challenging ❙ The dissemination of bilateral data should therefore be part of a longer term program and should focus on relevant corridors (pairs of countries with large flows)