This paper reviews the experience under the IMF's data standards initiatives and presents proposals for further refinements of the Special Data Dissemination Standard (SDDS) and the General Data Dissemination System (GDDS).

Subscribers have made significative progress in meeting the SDDS requirements since the second review of the SDDS at the end of 1998. This progress has been realized against the background of strong competing demands in national statistical agencies for resources to meet the Y2K challenge, and for European subscribers, launch the European Economic and Monetary Union. It is expected that in the first half of 2000, most subscribers will complete the remaining work to be in observance of the SDDS and that hyperlinks will be established from the Dissemination Standards Bulletin Board (DSBB) to the National Summary Data Pages (NSDPs) of most subscribers. Monitoring of observance is expected to begin by the end of June 2000. Thereafter, the staff will report periodically to the Board on the degree to which subscribers are adhering to the monitored elements of the Standard, and any significant issues (such as cases of egregious nonobservance) will be referred to the Executive Board.

At the March 23, 1999 meeting of the Executive Board, Directors approved the incorporation of the data template on international reserves and foreign currency liquidity into the SDDS as a prescribed component, with a transition period to run through the end of March 2000. Subscribers generally have had to strengthen their existing data collection systems for international reserves in order to compile the template data. Since June 1999, seven SDDS subscribers have begun disclosure of the template data. Most other SDDS subscribers appear to be on schedule to disseminate the template data by the end of the transition period. While a few countries have indicated that they would be in a position to disseminate the data more frequently and with a shorter lag than the current prescription for monthly data with up to a one-month lag, most are still working toward meeting the current prescription.

The staff is seeking Board approval for the use of the sample form for the reporting of template data and for the redissemination of country data on the Fund's external web site. The staff has invited countries to provide the template data to the Fund on a voluntary basis in this format, which entails the sample form (shown in Appendix I), providing some articulation of the template approved in March 2000. It seems clear that a more systematic approach to the compilation of the template data and the maintenance of a common database for these data by the Fund would support the Fund’s policy and surveillance activities and facilitate the redissemination of the data.

At the second review of the SDDS in December 1998, Directors generally supported the proposal to introduce a separate SDDS data category for external debt. Further, they requested the staff to return with precise proposals for the transition period for implementation of the new data category, after further consultation with countries, data users and other international organizations. During the staff consultations in 1999 and the February 2000 conference on capital flow and debt statistics, users generally, as well as compilers in many developing and transition countries, attached great importance to the need for timely and comprehensive external debt statistics. However, data compilers in many industrial countries and in some offshore financial centers doubted whether disseminating quarterly external debt statistics and data on forward amortization payments for their countries could be justified on the basis of their national statistical priorities and in light of the resource costs involved.

The staff proposes that the SDDS prescribe dissemination of quarterly external debt statistics within one quarter of the reference period, with information to be disseminated for the debt of the general government, the monetary authorities, the banking sector, and other sectors. Countries would also be encouraged to disseminate data on forward debt service obligations (with a split between amortization and interest) and on the domestic/foreign currency breakdown of external debt. A three-year transition period is proposed for implementing these proposals.

Phase I of the implementation of the GDDS focused on education and training. During FY1999 and FY2000, close cooperation with other international organizations was developed, especially with the World Bank for socio-demographic statistics. As a result, 55 countries have nominated a GDDS coordinator, and, so far, 34 of these countries have formally expressed an intention to participate in the GDDS. In implementing Phase II, STA would integrate work on the GDDS with its other services to member countries, while continuing to work closely with other interested international organizations. The staff plans to begin posting GDDS metadata (including plans for improvement) on the DSBB as soon as a critical mass of countries has approved its metadata for dissemination. Also, the staff recommends that selected components of external debt—namely public and publicly guaranteed external debt and the associated debt service schedules—be upgraded from an encouraged to a core framework (and data category) of the GDDS.

Several recent initiatives draw on the data dissemination standards. In the IMF’s work on standards and codes, the experimental reports on their observance have involved both the SDDS and the GDDS. The proposal of the Basle Committee on Banking Supervision for a New Capital Adequacy Framework and the IMF's Contingent Credit Line are initiatives where the SDDS has been placed in a benchmark role.

An in-depth study of the DSBB was undertaken by consultants in 1999. Improvements to the DSBB, based on the consultant’s recommendations, will be introduced in due course. The staff believes that, to ensure the successful future development of the DSBB, it is imperative to further strengthen its links with actual data beyond those encompassed by hyperlinks to NSDPs. The staff is proposing to explore the possibility of reaching agreements with commercial vendors for hyperlinks from the DSBB to commercial sites that would disseminate SDDS data.

I. Introduction and Scope of the Review

1. The Executive Board established the Special Data Dissemination Standard (SDDS) in March 1996 and the General Data Dissemination System (GDDS) in December 1997. The SDDS aims to guide countries that have or seek access to international financial markets in the dissemination of economic and financial data to the public. The GDDS, the second tier of the Fund’s data standards initiatives, emphasizes the development of sound statistical systems as the basis for the timely dissemination of data to the public.



2. In establishing the SDDS, Directors emphasized that it should be implemented flexibly to adapt to meet new challenges and changing circumstances and called for reviews of the SDDS to provide opportunities to make needed adjustments. The first review took place in December 1997 and the second review in December 1998, with further consideration in March 1999. On the occasion of the second review, Directors took actions to further strengthen the SDDS and decided on a third review. The conclusions of these discussions are summarized in Box 1.



3. In establishing the GDDS, Directors decided that it would be implemented in two phases, with Phase I, spanning FY1999 and FY2000, focused on training and information. The first phase consisted of regional seminars and the development of pilot metadata. In the second, or implementation, phase, the staff will integrate the GDDS into the work of the Statistics Department (STA) and disseminate metadata on the Dissemination Standards Bulletin Board (DSBB).



4. This paper provides input for the third review of the SDDS and updates developments in the GDDS. Section II reviews recent experience of the SDDS, including the strengthened international reserves data category and makes proposals for the implementation of the new external debt data category. Section III reports on the wind-up of the first phase of the GDDS and outlines a work program for the second phase. Section IV examines trends affecting the Fund's data standards initiatives and seeks guidance on possible paths for future developments. Section V deals with staff resource issues, and Section VI provides issues for discussion.




 

Box 1. Board Discussions on the SDDS and GDDS since their Establishment: Key Conclusions The First Review of the SDDS (December 1997) Endorsed proposed procedures for modifying the SDDS. Agreed that consideration be given to modifying the international reserves data category. Decided that a timetable for implementing the international investment position (IIP) be agreed upon at the second review. Agreed with preliminary proposals for dealing with nonobservance after the end of the transition period. Agreed that a refined proposal be discussed at second review. Decided that the DSBB remain free to users. Established the GDDS. Data Availability, Dissemination, and Provision to the Fund (September 1998) Asked staff to develop data template and guidelines on international reserves. Endorsed inclusion of data on short-term foreign currency debt of central government, with same periodicity and timeliness as international reserves. Agreed to aim for weekly data on reserves with one-week lag. Agreed that hyperlinks should be prescribed. Identified improvements in external debt data as a high priority. Endorsed suggestions to look at ways to enhance the DSBB. Second Review of the SDDS (December 1998) Endorsed proposed additional temporary flexibility option for SDDS subscribers through the end of 1999. Differed in views on proposed template for the disclosure of reserves data. Agreed to revisit in early 1999 the issue on the coverage of the template, and the periodicity and timeliness of data dissemination. Generally supported proposed introduction of a separate SDDS data category for external debt to be disseminated quarterly with one-quarter timeliness. Asked staff to return with a detailed proposal for transition periods. Supported a prescribed specification of a three-year transition period for dissemination of annual IIP data with a six-month lag. Agreed to revisit in third review of SDDS the issue of monitoring of observance by staff. Endorsed proposed inclusion of hyperlinks as required feature of the SDDS, with one-year transition period. Supported suggested modifications to the GDDS consistent with proposed enhancements to SDDS, but decided to wait for finalization of SDDS adjustments before approving changes to GDDS. Second Review of the SDDS – Further Considerations on Data Template on International Reserves and Foreign Currency Liquidity (March 1999) Adopted SDDS prescription for dissemination of full data for template on a monthly basis, with a lag of no more than one month. Encouraged dissemination on a weekly basis, with a one-week lag. Agreed on transition period for observance through March 31, 2000. Agreed to reassess prescriptions for periodicity and timeliness in context of third review of SDDS. Macroprudential Indicators (MPIs) and Data Dissemination – The Role of the Fund (January 2000) Endorsed conclusions of September 1999 Consultative Meeting that the Fund undertake a survey on the availability of data and on dissemination and compilation practices relating to MPIs. Underscored the distinction between prudential data provided to the Fund in the context of bilateral surveillance and MPIs compiled with a view toward their dissemination to inform the public. Supported collaboration of the Fund with other international organizations to avoid duplication of efforts and to work toward developments of incentives for national authorities to disseminate MPIs. Supported work by staff on development of a core set of MPIs. Endorsed strategy of working now with unharmonized national data, but striving to develop internationally harmonized data over the medium term. Considered that inclusion of MPIs in SDDS is premature. Endorsed the publication of the background paper on MPIs.

 

II. The Special Data Dissemination Standard

5. Background on the development of and experience with the SDDS is summarized in Box 2. Consultation on the strengthening of the SDDS has been a continuous process since its inception (see Box 3). Progress has been achieved in implementing the data template on international reserves and foreign currency liquidity. Initiatives to improve access to external debt data are on schedule. A paper on so-called macroprudential indicators (MPIs) provided the basis of a Board discussion in January 2000. Staff are following up on the Board's guidance, as summarized in Box 1, and there are no new developments that need to be addressed in this paper.




A. Data Template on International Reserves and Foreign Currency Liquidity

6. At the March 23, 1999 meeting of the Executive Board on the Second Review of the SDDS—Further Considerations, Directors approved the incorporation of the data template on international reserves and foreign currency liquidity into the SDDS as a prescribed component. The template had been prepared by the staff in cooperation with the Committee on the Global Financial System of the G-10 group of central banks. Directors established a transition period for compliance with the SDDS requirements of the template to run through the end of March 2000.



7. In taking the decision, Directors asked the staff to prepare operational guidelines to assist member countries in completing the template and to consult with member countries on the guidelines. A first draft of the Operational Guidelines for the Data Template on International Reserves and Foreign Currency Liquidity was ready in May 1999. Following discussions with a number of member countries and in international fora, revised provisional guidelines were issued in October 1999. The guidelines were sent to Fund Governors of all member countries under cover of a letter from the Managing Director dated October 28, 1999. Executive Directors were briefed on the guidelines at an informal meeting of the Executive Board on November 8, 1999.




 

Box 2. Background on SDDS Development and Experience 1. The origins of the Special Data Dissemination Standard (SDDS) are rooted in events in 1994/95 that underscored the role that information deficiencies could play in contributing to market turmoil. The SDDS was envisaged as providing on a voluntary basis a set of data dissemination standards, representing an effort to codify existing good practice, to which countries participating in international capital markets, or aspiring to do so, could subscribe. Subscription to the SDDS is voluntary; however, observance of the Standard by subscribers is mandatory. 2. After wide-ranging consultations by staff with users and producers of data and feedback from the Executive Board, the SDDS was established in March 1996, encompassing four key dimensions of statistical practice: data (within which coverage, periodicity and timeliness), access, integrity, and quality. Under the SDDS, subscribing countries would provide to the public more comprehensive, reliable, accessible, and timely economic and financial information. The SDDS was not expected to prevent economic crises. Nevertheless, the disturbances that began in Asia in 1997 have further underscored the role that information deficiencies can play and have indicated the need for SDDS refinements. In particular, enhancements concerning the specifications for international reserves and external debt were seen as essential. 3. The international community is now engaged in an effort to improve the architecture of the international monetary system. Much of this work involves translating to the international domain standards and codes of practice that help to support efficient markets within individual countries. The areas where such work is underway include accounting standards, bankruptcy procedures, and fiscal and monetary transparency. The SDDS can be seen as an early undertaking by the Fund in these standard-setting activities. As part of the evolving SDDS process, the Fund has undertaken wide-ranging consultations with countries and other international agencies, as well as its first experiments with public discussion papers of Fund policy proposals and the use of an Internet web site to communicate information on data standards. All of these provide useful lessons for future activities in the area of standards. 4. The SDDS has stimulated a large volume of data dissemination activity over the 4 years since its inception. By March 9, 2000, forty-seven countries were subscribers, including major industrial countries and many emerging market economies. Given the target audience—countries participating in international capital markets or aspiring to do so—this represents a high subscription rate, consistent with expectations when the SDDS was first established. Encouraged by the SDDS, subscribers have introduced wide-ranging improvements in statistical practices, to the benefit of the countries and the international community. In September 1996, the Fund opened the DSBB on the World Wide Web. This electronic bulletin board provides easily accessible information about the statistical practices, or "metadata," of subscribers, available at the click of a mouse. 5. Early on, it was clear that users of the DSBB would find it more useful if there were a direct link to actual economic and financial data, in addition to the information it provided about data practices. To that end, hyperlinks from the DSBB to country web sites—with the data on the country site to correspond to the data described on the DSBB—were introduced. By March 9, 2000, hyperlinks to country data sites were in place for 19 subscribers and were under review by the staff for another 23 subscribers. Five subscribers have web sites on which some or all SDDS data categories are disseminated, but progress on the development of NSDP is unknown. Usage of the DSBB has surged over time, from some 25,000 hits per month in September 1996 to more than 190,000 per month in February 2000 (see Figure 1). The goal for the period ahead is the further enhancement of the DSBB as a user-friendly way to access information concerning the SDDS, the statistical practices of subscribers, and actual economic and financial data for subscribers.

 



 

Box 3. Consultation on Strengthening the SDDS Subsequent to the design and implementation of the SDDS during the 1995–97 period, the staff undertook the following consultations on the strengthening of the SDDS: In the wake of the Asian crisis, efforts began to strengthen the SDDS in the areas of international reserves and external debt. The issues have been discussed in Board papers, in regular contacts with country officials, and with other official agencies. In addition, in order to solicit feedback from data users and compilers, a consultation paper on SDDS proposals concerning international reserves was placed on the Internet in May 1998. The responses provided input to finalize the staff’s proposals. A consultation paper on external debt, seeking the views of compilers and users, was placed on the Internet for reaction in October 1998. The consultation elicited valuable information that was used in the preparation of the Board paper for the December 1998 Second Review of the SDDS. A second consultation paper on external debt, seeking more specific information on various components of external debt and on country compilation and dissemination practices, was sent to compilers in the SDDS-subscribing countries in July 1999. The paper was also posted on the Internet. A consultative meeting on MPIs was organized by the Fund’s Monetary and Exchange Affairs Department and STA in September 1999. The meeting provided Fund staff with an opportunity to learn from the experience of the public and private sectors with specific types of MPIs and to discuss the MPIs that could be used most effectively by the Fund in its surveillance work (see also Box 1). A conference—conference on capital flow and debt statistics: Can We Get Better Data Faster?—was hosted by the Fund on February 23–24, 2000, in cooperation with the Working Group on Capital Flows of the Financial Stability Forum. The conference brought together some 120 senior-level data users in the official and private sectors and compilers of statistics. A primary objective of the conference was to identify data requirements in connection with capital flow and debt statistics and what actions and resources would be required to provide better data on a more timely basis.

 

8. Recognizing that compiling the template data would be a significant undertaking for some countries, the Managing Director, in the cover letter to SDDS-subscribing countries, asked these countries to provide a frank indication of whether they could meet the requirements of the template by the end of the transition period.¹ The letter also noted that the staff would provide technical assistance, if needed.




Progress in implementing the data template

9. Since June 1999, seven SDDS subscribers have begun disclosure of international reserves data based on the template (viz., Canada, France, Germany, the Netherlands, Switzerland, the United Kingdom, and the United States).



• Except for the United States, these countries disseminate the data on a monthly basis with up to a one-month lag; the United States disseminates its data on a weekly basis with a four-day lag.



• The data are disseminated in national media, including on the Internet web sites of countries’ central banks and ministries of finance.



• The presentation of the data varies among the countries. Some follow the structure of the data template; others modify their existing national presentations on international reserves to incorporate the template data.



• Countries note that they will refine their coverage in coming months as they improve their data collection methods.



10. In response to the Managing Director’s query on implementation of the template, at time of writing, 32 countries, in addition to those already disseminating the template data, had indicated that they would disseminate the data by the end of transition period. With very few exceptions, countries also expressed support for the data template initiative insofar as it would strengthen the international financial architecture.



11. In assisting countries in implementing the template, the staff has responded to inquiries from 28 countries seeking practical guidance in compiling the template data. Most inquiries have been technical and country-specific.² The staff has also canvassed selected countries in various regions of the world to learn what specific steps, if any, they need to take and whether they foresee major difficulties. Although the complexity of reporting can vary from country to country, depending on the nature, size, and intricacy of the financial transactions the authorities undertake, the canvassing results show that:



• Existing data collection systems generally need to be strengthened to provide the greater detail on a more frequent basis that is called for by the template.³



• Coordination mechanisms need to be established to gather data from various sources to meet the coverage specified in the template; this is particularly the case for foreign currency assets and liabilities of the central government, which usually encompasses numerous agencies.



• Accounting practices used in the data-providing agencies need to be reconciled with the underlying concepts of the template; this is viewed as a major technical difficulty.



12. The staff expects that most current SDDS subscribers will begin disseminating the template data on schedule. Moreover, the quality of the data is expected to improve over time as subscribers gain experience. The staff plans to produce a revised version of the Operational Guidelines for the Data Template on International Reserves and Foreign Currency Liquidity after the end of the year. The revised version will take into account countries’ experiences in implementing the data template.



13. At the Executive Board meeting of March 23, 1999, Directors agreed that the prescriptions for timeliness of reserves data dissemination should be reassessed in the context of the third review of the SDDS, bearing in mind the advantages of more frequent and timely disclosure, as well as the costs of disclosure. As noted above, it is clear that countries will need some time to gain experience in operating with their new template data systems before comprehensive data can be compiled on a regular and timely basis. In the staff’s contacts with SDDS-subscribing countries, few countries have indicated that they would be in a position to provide template data on a weekly basis with a one-week lag in the near term, but most are still working toward meeting the current prescriptions. On the basis of the foregoing, the staff proposes that the existing prescriptions for timeliness and periodicity not be modified at this time. Directors may wish to reexamine the issue at the time of the next review of the SDDS.




Development of a Fund database on the reserves template

14. The staff informed the Executive Board of its initiative to establish a common database at the Fund for country data on the template on international reserves and foreign currency liquidity (SM/99/312,12/27/99). Such a database will provide a single source for the data for use by staff to support surveillance and policy development. A common database presented in a common format and in a common currency would also enhance comparability of the template data across countries. In addition, the availability of a common database would enable the Fund to redisseminate the data to the public, thereby facilitating access to the data by market participants and other users and fostering greater transparency.



15. This past January, Fund staff invited SDDS subscribers to cooperate with the Fund, on a voluntary basis, to provide their reserves template data to the Fund using a common data format, at the same time that they disseminate the data to the public. To ensure timely entry into the Fund database, countries were asked to provide the template data via email or other electronic means. In order not to compromise timeliness, the country data would be entered into the Fund database without prior review by Fund staff.



16. The common data format that the staff proposes to be used in the Fund database for the reserves template data is that of the sample form shown in Appendix II of the Operational Guidelines for the Data Template on International Reserves and Foreign Currency Liquidity (reproduced in Appendix I of this paper). The sample form was designed to facilitate data compilation by countries, as well as to promote data comparability across countries. The sample form presents all items of the template, including those called for in the footnotes of the template. Where the template does not specify coverage for some general data categories, the sample form provides specific items for such general data categories in order to promote uniformity of data reporting by countries. The data coverage in the sample form is consistent with that of the data template. Directors may wish to note that a number of countries, including those in the Eurosystem, plan to use the sample form to disseminate template data. The common currency that the staff proposes to be used in the Fund database, in addition to the currency used by countries in reporting the template data, is the U.S. dollar.



17. The staff believes that it would be useful for the Fund to redisseminate the template data in the Fund database through the Internet, taking advantage of the timeliness of this medium. It is expected that the data could be redisseminated in the Fund’s external web site within 24 hours after receipt of the data by the Fund. To avoid giving the impression that the Fund has endorsed the reliability of the redisseminated data, a disclaimer would be included. The staff’s objective is to have the Fund database and the Internet site operational by March 31, 2000 to coincide with the end of the transition period for the dissemination of the template data under the SDDS.



18. The staff is seeking Board approval of the sample form, which represents some articulation of the reserves data template, as the format for dissemination of the template data by SDDS subscribing countries and as the format for the transmittal of the country data to the Fund database. The staff is also seeking the Board's concurrence in the redissemination of the country template data submitted to the Fund on the Internet.




B. External Debt Data Category

Introduction

19. At the December 2, 1998 meeting on the second review of the SDDS, Directors generally supported the staff’s proposal to introduce a separate data category for external debt with data for three main components—general government, the monetary authorities and banks, and nonfinancial public corporations and the private sector—and with further breakdowns, including by maturity, to be disseminated quarterly with one-quarter timeliness.⁴ Several Directors, however, considered that the staff proposal to include forward-looking data on external debt service was too demanding, especially for private sector obligations.



20. Directors requested the staff to return with precise proposals for the transition periods for implementation of the proposals on external debt in the SDDS after further consultation with countries, data users, and other international organizations. The staff has conducted these consultations, during which the question of whether to include forward-looking data on debt service obligations was also considered. The proposed coverage of the external debt category and the transition periods for implementation are discussed later in this section.



21. In the past year, the need for improved external debt data has been reinforced in other fora. At its inaugural meeting last December, the G-20 was supportive of a proposal for countries to develop more comprehensive indicators of vulnerability, including the debt structures of the public and private sector. In its Communiqué of September 26, 1999, the Interim Committee encouraged further work by the Fund on the SDDS, including on strengthening external debt data. Two international initiatives underway to improve access to and understanding of data on external debt are described in Box 4.




 

Box 4. International Initiatives to Improve Access to and Understanding of External Debt Statistics 1. Two important initiatives to improve access to and understanding of data on external debt were launched in 1998 in connection with the work program of the Inter-Agency Task Force on Finance Statistics (TFFS).1 First, beginning in March 1999, a new series of quarterly releases of statistics on external debt for 176 developing and transition countries is being jointly disseminated by the Bank for International Settlements (BIS), the Fund, the Organisation for Economic Co-operation and Development (OECD), and the World Bank. From the start, it was understood that these data, largely constructed from creditor and market sources of information, would not provide a completely comprehensive and consistent measure of a country’s total external debt. For example, they do not cover large elements of trade credit, debt owed to foreign direct investors, nonresident deposits held with domestic financial institutions, and private placements of securities, as well as domestically issued debt securities held by nonresidents. 2. Nevertheless, the joint BIS-IMF-OECD-World Bank statistics bring together for the first time the best and most timely internationally comparative data currently available on external debt. This initiative, while highly desirable, is not a substitute for the important task of improving external debt data compiled from national debtor sources. This message was strongly reinforced by participants at the conference on capital flow and debt statistics: Can We Get Better Data Faster? (February 23–24, 2000), which was sponsored by the Fund in cooperation with the Financial Stability Forum's Working Group on Capital Flows. 3. Second, to aid in the task of improving national-source data, the TFFS commenced work on External Debt Statistics: Guide for Compilers and Users (Debt Guide). A preliminary draft of core chapters was reviewed by the TFFS in January 2000. The Debt Guide, which updates a 1988 inter-agency publication, will provide international statistical standards for the measurement of external debt as well as guidance in the analytical use of the data and the sources and methods for their compilation. The target date for publication, following extensive consultations with compilers and users, is June 2001. A draft of the core chapters will be available for testing in a series of workshops and seminars. The first seminar is being held at the IMF-Singapore Regional Training Institute on March 21–24, 2000. It is being presented collaboratively by the Fund, the Australian Bureau of Statistics, and the Singapore Department of Statistics and will include about 30 participants from 13 Asian economies and the Asian Development Bank. __________________________ 1 The recent work of Fund-chaired Inter-Agency TFFS has included representatives from the BIS, the Commonwealth Secretariat, the ECB, Eurostat, the IMF, the OECD, the Paris Club Secretariat, UNCTAD, and the World Bank.

 

22. The staff consulted widely on the appropriate transition periods for implementation of the new proposals for external debt data in the SDDS. A consultation paper on external debt that contained a detailed questionnaire on external debt compilation and dissemination practices was sent to the SDDS coordinators in mid-1999, and responses were received from national compilers in all 47 of the SDDS-subscribing countries. The questionnaire was also posted on the DSBB for the benefit of other countries and data users. The results of the debt questionnaire were put before the IMF Committee on Balance of Payments Statistics and the TFFS, and consultations were also undertaken with selected users and international organizations.



23. Most recently, the staff presented the results of the debt consultations at last month’s conference on capital flow and debt statistics: Can We Get Better Data Faster? The conference provided a unique opportunity for policymakers, data users in the private sector, and statistical compilers to exchange views on what are the most pressing requirements for data on capital flows and external debt and how any new requirements can be met.



• Policymakers and data users stressed the need for improved data on external debt as well as supplementary information related to these statistics, such as a forward amortization schedule and information on the currency composition of debt.



• Data compilers from many of the industrial countries and in some offshore financial centers viewed dissemination of quarterly external debt statistics and data on forward amortization payments for their countries as a low priority. This reflected national statistical priorities as well as concerns regarding resources, compilation difficulties, and the burden on reporters that the new collections would entail.



24. Thus, there emerged a view that national statistical priorities often differed from international priorities driven by the data needs of policymakers and market participants, which needs to be borne in mind in reviewing the SDDS.




Responses to consultation on external debt statistics

25. The consultation focused on two important elements of external debt data—quarterly periodicity and timeliness, and information on forward-looking debt service obligations.⁵ In various fora, users, market participants, and policymakers stressed the need for timely and comprehensive data on external debt, debt service schedules, and the currency composition of debt. Some asserted that the costs to the international community as a whole from not having such data would be far in excess of the costs of producing them.



26. The compilers' responses to the external debt questionnaire showed that 70 percent of the developing/emerging market economies that subscribed to the SDDS currently compile and disseminate either monthly or quarterly external debt and/or international investment position (IIP) statistics.⁶ Only 35 percent of the industrial countries disseminated monthly or quarterly data in either of these areas. The coverage of the statistics varied. The industrial countries that did not disseminate monthly or quarterly external debt/IIP data mostly disseminated annual IIP statistics, but they indicated that some external debt data were available on a more frequent basis, mainly in respect of the banking sector and the monetary authorities.⁷



27. Those countries that do not disseminate monthly or quarterly external debt/IIP data provided a range of responses regarding required transition periods for their countries: around a third indicated a year or less would be required and another third indicated a period of between one and three years. Most of the other countries did not specify a suitable transition period for this work. Several of the countries that said they disseminated quarterly debt information also indicated that they required a short transition period. Many of the industrial countries stressed that, in order to compile quarterly external debt statistics, some data components would have to be estimated.⁸



28. The need for countries to strengthen their ability to assess their external vulnerability points to the need for a debt service schedule for external liabilities in order to identify obligations that are due to be paid in the short term. Such information is also useful for creditors to monitor countries’ foreign obligations. The responses to the debt questionnaire revealed that only about 40 percent of the countries were in a position to compile elements of an amortization schedule, mostly in respect of foreign liabilities in the form of international bonds and loans. A greater number of countries (60 percent) were able to compile debt data on a residual maturity basis.⁹ In both cases, compilers frequently could not provide maturity information for the full range of debt components, e.g., data on trade credits, deposits, and nonequity liabilities to foreign direct investors were frequently not available. Overall, compilers in developing/emerging market economies reported better coverage.



29. Many of the compilers from the industrial countries saw little need for an amortization schedule for their countries. In general, many compilers from developing/ emerging market countries expressed the opposite view. Overall, compilers had difficulties in specifying transition periods for developing such data. They also indicated that the development of such data would entail substantial changes to national compilation systems, which would require resources and entail higher respondent burden. The latter view was strongly reinforced by the IMF Committee on Balance of Payments Statistics.



30. At the time of the second review of the SDDS, Directors generally thought that it was important to show a sectoral presentation of external debt, and this point also has been made forcefully by data users. Compilers preferred a sectoral presentation over the functional presentation (direct investment, portfolio investment, and other investment) emphasized in the Balance of Payments Manual, Fifth edition (BPM5). In the debt questionnaire, 58 percent favored the sectoral presentation and only 12 percent preferred the functional classification. The remaining countries expressed no preference, as their debt compilation systems could generate data for both classifications or they saw value in disseminating data in both bases.



31. Compilers identified a number of difficulties in compiling comprehensive external debt statistics, such as coverage of the external obligations of the private sector, measuring short-term trade credit, determining residency of holders of domestically issued debt securities, and recording repurchase agreements. Many of these issues also pose difficulties in compiling comprehensive balance of payments and IIP statistics. It is recognized that in seeking timely external debt data, there will be a trade-off between data quality and timeliness. The need to provide adequate resources to meet the new data demands must be recognized by policymakers, a theme stressed by compilers at the conference on capital flow and debt statistics.



32. In the consultation, some countries urged the Fund to recognize the different circumstances, priorities, and constraints faced by different economies. This theme also emerged in the conference on capital flow and debt statistics.




Staff proposals on external debt

33. The staff’s proposals attempt to strike a balance that recognizes the need for more timely information on external debt and the difficulties faced by compilers in producing data that are of acceptable quality. The staff notes that the successful implementation of the new SDDS data category for external debt statistics will require the support of all countries—industrial and developing—including that of their compilers.¹⁰



34. The staff proposes that the SDDS prescribe dissemination of quarterly external debt statistics within one quarter of the reference period, with information to be disseminated for the debt of the general government, the monetary authorities, the banking sector, and other sectors.¹¹ Data should also be broken down by maturity—short-term and long-term—on an original maturity basis and by instrument, as set out in the BPM5.¹²



35. It is envisaged that, to the extent feasible, the new SDDS data category on external debt would be consistent with the framework for IIP statistics presented in the BPM5, which facilitates compilation efforts and ensures consistency with other statistical systems. External debt relates to liabilities to nonresidents regardless of currency of denomination of the debt.¹³



36. The recommendation to report maturity on an original basis reflects two main considerations. First, the present international statistical standards for the balance of payments, the IIP, and the national accounts are based on this concept of maturity. To prescribe another concept, such as residual maturity, would break the link with other international data sets. In addition, given the SDDS prescription for dissemination of the IIP, a different prescription would create considerable additional compilation costs as two different measures of external liabilities would be prescribed. Compilers emphasized this point in the consultation. Second, from an analytical viewpoint, the quarterly dissemination of position data on an original maturity basis for countries that have—or might have—ready- access to the international capital markets would provide frequent observations of the extent to which member countries’ rely on external short-term finance as opposed to longer-term finance, and how this reliance is changing over time ¾ an important consideration in external vulnerability analysis.



37. Nonetheless, a measure of short-term indebtedness based on original maturity would not reveal that component of long-term debt that is to be paid in full (at maturity) or in part (prior to maturity) within a period of one year. Thus, in addition to a position statement of external debt based on the original maturity concept, the staff proposes that the SDDS encourage the dissemination of supplementary information on prospective debt service obligations, in which the principal and interest components are separately identified twice yearly for four quarters and two semesters ahead, with a lag of one quarter. The data should also be broken down into sectors—general government, monetary authorities, the banking sector, and other sectors.



38. Fund staff considers that the dissemination of a debt service schedule should be considered as best practice that promotes greater transparency and understanding of a country’s external debt situation. However, the staff has recommended a debt service schedule on an encouraged rather than on a prescribed basis owing to the results of the questionnaire: compilers in many of the advanced industrial countries doubted the value of such a schedule, given the ability of most residents to readily refinance maturing debt and, as a consequence, questioned the value of employing resources for compiling such data.



39. A second concern raised by compilers in the industrial countries and in offshore financial centers is the need to disseminate information on external financial assets in order to be able to fully evaluate a country’s external debt position. In this regard, the SDDS encourages the dissemination of quarterly IIP statistics, which include data on external financial assets, as well as external liabilities. If compilation of a quarterly IIP statement is not feasible, compilers may wish to consider highlighting, by way of supplementary information, available information on selected foreign assets, such as those of the banking sector. The latter has special relevance to the operations of offshore financial centers.



40. Participants at the recent conference on capital flow and debt statistics stressed the importance of a domestic/foreign currency breakdown of debt, which may be of help in assessing vulnerabilities in connection with exchange rate changes. However, it is recognized that the use of financial derivatives to hedge exchange rate risk can significantly alter the effective currency composition of short-term debt and limit the usefulness of information on the domestic/foreign currency breakdown. In the debt questionnaire, 42 percent of the respondents indicated that it was very important for countries to disseminate information on the currency composition of external debt while another 42 percent said that it was somewhat important to disseminate such data. Two-thirds of the respondents said that data were available either for individual currencies or a domestic/foreign currency split of external debt. Other countries had currency detail for selected sectors. The staff would propose that the SDDS encourage countries to disseminate a domestic-foreign currency breakdown of external debt each quarter, within one quarter of the reference period.



41. In view of the foregoing, it is proposed that the transition period for disseminating quarterly external debt statistics end on March 31, 2003, fifteen months after the end of the transition period for IIP statistics. The staff believes that countries should not be permitted flexibility with respect to quarterly periodicity and timeliness for the external debt data category, because such data, like international reserves, are essential in assessing external vulnerability.¹⁴



42. The efforts of countries to develop their external debt statistics will be supported by the TFFS through the Debt Guide, through a series of regional seminars and workshops, and through the ongoing technical assistance programs of the agencies participating in the work of the TFFS.



43. The Fund would encourage subscribers, meanwhile, to disseminate quarterly external debt statistics, as appropriate and feasible, and to report these data to STA, which will begin work to develop a database on debtor-source statistics in support of the Fund's surveillance work and for dissemination to the public.



44. Under the second review of the SDDS, Directors approved a three-year transition period ending on December 31, 2001 for the dissemination of annual IIP data as a prescribed SDDS data category. These data are to be disseminated within two quarters of the end of the reference year. In the course of the debt consultation, a substantial number of subscribers indicated that the timeliness requirement of six months would be difficult to achieve. Provided that quarterly external debt data are disseminated with a one-quarter lag, the Fund staff would propose an extension, to nine months, in the timeliness requirement for dissemination of annual IIP statistics.




A. Observance and Uses of the SDDS

Observance and monitoring

45. The second review of the SDDS in December 1998 provided for staff monitoring of the data dimension—coverage, periodicity, and timeliness of the data—and the advance release calendars (ARCs) of the access dimension, and established that monitoring would become fully effective when hyperlinks had been established between the DSBB and NSDPs for all subscribers. Other elements of the SDDS are on a self-declaration basis by subscribers.



46. This section provides a brief status report on observance and monitoring of the SDDS as of March 9, 2000. It summarizes the work that remains to be done in bringing subscribers' practices into line with the Standard. It also summarizes developments in establishing hyperlinks between the DSBB and NSDPs and in posting summary methodology statements on the DSBB. In addition, the section reviews the approach for addressing observance issues and the work initiated subsequently by the staff for monitoring observance.




Observance status

47. The total number of subscribers to the SDDS has not changed since the second review and stands at 47. The staff is working with seven additional member countries that have expressed an interest in subscription. Moreover, developments since the second review aimed at enhancing transparency within the new information architecture may provide incentives to nonsubscribers who participate in international financial markets to consider subscription more actively (see below).¹⁵



48. As of March 9, 2000, thirteen SDDS subscribers had provided evidence that they were in observance of the SDDS specifications of the data dimension and ARC elements of the access dimension, and there remained a total of 58 transition plans outstanding among those subscribers not yet in observance (see Appendix II). Nevertheless, there has been substantial progress among all subscribers over the course of 1999 and the beginning of 2000 in meeting the requirements of the SDDS (see Figure 2). This progress has occurred against the background of competing demands for resources within statistical agencies, especially to meet the challenges posed by the transition to a new millennium (Y2K problems) and by regionally based initiatives—notably implementation of Stage Three of the European Economic and Monetary Union (EMU). Moreover, although the SDDS was aimed at economies with developed statistical systems, for some economies, the SDDS also provided a framework for the introduction of new compilation systems—such as quarterly national accounts and balance of payments—and, as a result, meeting the Standard has taken longer than expected.

     

    

     

49. In addition to the metadata base pages that describe the data disseminated under the SDDS, subscribers are required to provide summary methodology statements that describe the statistical compilation practices for each data series in some detail. These statements contain information that users may access to determine if the data disseminated are of sufficient quality for their purposes. As of March 9, 2000, the staff has posted a total of 167 summary methodology statements for 38 countries and work is underway to substantially increase this number in the months ahead.




National Summary Data Pages

50. Under the second review of the SDDS, the Executive Board made the establishment of a hyperlink from the DSBB to the NSDP mandatory for each subscriber. This decision provided a unique tool to enable monitoring of the SDDS by the staff. Although the staff was to begin monitoring of the data dimension and the ARCs during 1999, monitoring was to become fully effective when all hyperlinks to NSDPs were in place.



51. A total of 42 SDDS subscribers have established NSDP web sites on which all the SDDS data categories are disseminated. Hyperlinks from the DSBB to these web sites have been established for 19 subscribers, and the remaining 23 draft NSDPs are under review by the staff. The five subscribers without NSDPs do have web sites on which some or all of the SDDS data categories are disseminated, but there is no single web site for the dissemination of all SDDS data categories for these subscribers.



52. Subscribers have encountered a broad range of difficulties in establishing NSDPs during the past year, including resource constraints and institutional rigidities that hampered effective coordination among the various statistical agencies responsible for disseminating SDDS data. Resource constraints have been associated with Y2K problems as national statistical offices were required to give highest priority in their use of information technology (IT) resources to addressing these problems. As a result, work on establishing an NSDP often was delayed until the end of the year. A consequence of this sequence of events has been that the large number of NSDPs that became available for review during the closing months of 1999 overwhelmed the Fund staff resources available for review and processing. It is expected, therefore, that a substantial number of hyperlinks to NSDPs will be established in the first half of 2000.




Monitoring observance of the data dimension and the Advance Release Calendar

53. The staff has begun monitoring observance of the data dimension and the ARC of the access dimension on a pilot basis for those subscribers that have met the SDDS requirements and have hyperlinks from the DSBB to NSDPs. Notwithstanding the developments noted above, structured monitoring of the Standard must begin as soon as practicable after the completion of this review. The absence of hyperlinks to NSDPs for some subscribers will make monitoring more difficult and time-consuming than was envisaged at the time of the second review because, for these subscribers, several web sites may need to be checked to ensure as complete coverage as possible of the SDDS data categories (see below). Nevertheless, beginning monitoring of the SDDS, with accompanying notifications to subscribers who do not meet the Standard, should provide incentive to subscribers to complete outstanding work on SDDS issues as soon as possible. Instances of egregious nonobservance, once identified and for which the usual channels for resolving such instances had been exhausted, could lead to notification of the Executive Board.



54. The second review of the SDDS also provided a framework for dealing with nonobservance of the Standard (see Box 5). The approach put forward sought to achieve two objectives, namely (1) maintaining the credibility of the metadata posted on the DSBB and (2) maintaining a high level of transparency to ensure that users of the DSBB are aware of the steps involved.




 

Box 5. Monitoring of Observance and Dealing with Nonobservance In the summing-up by the Acting Chairman (EBM 98/131, 12/21/98), Executive Directors underscored the importance of ensuring that subscribers meet their commitments under the SDDS in order to maintain the credibility of the Standard. Most directors agreed that a structured approach to monitoring of observance by the staff was desirable. Many directors emphasized the need for a transparent approach to dealings with instances of nonobservance of commitments under the SDDS, which would ensure that the users of information were alerted to the existence of problems at an early stage. The Acting Chairman concluded that while most Directors could go along with the proposals put forth for addressing nonobservance (see below), this was an area in which the Fund’s policies would need to evolve in the light of experience. The Executive Board paper SM/99/65 (5/11/99) contains the proposals on observance and monitoring adopted by the Executive Board on December 21, 1998. These proposals may be summarized as follows: Fund staff would continuously monitor the data dimension and the advance release calendars element of the access dimension of the SDDS. Fund staff would try to resolve each case of nonobservance directly with the subscriber, and, if necessary through the Executive Director. If no resolution ensued, the matter would be brought to the attention of the subscriber’s Governor for the Fund and the Fund could post a notice on the DSBB describing the problem, the subscriber's response, and efforts underway to remedy it. If no remedy resulted, the matter would be referred to the Executive Board, which could decide that the subscriber was not in observance of the SDDS and (a) place a notice to that effect on the DSBB and, if satisfactory corrective measures were not taken, (b) could decide to remove the metadata of the subscriber from the DSBB.

 

55. It has not been possible to gain experience with the procedures agreed by the Executive Board for addressing nonobservance because continuous structured monitoring of observance has not been possible. Nevertheless, the staff believes that in cases of egregious nonobservance, the removal of a subscriber's metadata from the DSBB—one step that is open to the Board—could undermine the Fund's efforts to promote transparency by depriving market participants of information on how a subscriber was not meeting the Standard, as well as metadata for other data categories that were in observance of the SDDS. In these cases, a notice on the DSBB explaining that a subscriber was not in observance for a specific data category should provide a powerful incentive for the subscriber to come into observance as soon as possible.




Technical Cooperation Action Plans/Technical Assistance link and the role of evaluations

56. As part of the effort to integrate more fully technical assistance (TA) with surveillance and program activities, in June 1999 the Fund commenced a pilot program involving the preparation of Technical Cooperation Action Plans (TCAP). The TCAP is planned to underpin improvements in economic and structural policies and comprises elements covering fiscal and monetary policies and statistics. The TCAP defines TA objectives and describes the anticipated arrangements for implementation, monitoring, and assessment of TA efforts. As such, it can contribute to the development and evaluation of a country’s preparedness to adhere to internationally recognized standards, including the SDDS. In the cases of Brazil, the Russian Federation, and Uruguay, STA missions have formulated, in close collaboration with the authorities, detailed plans of action to guide these members to reach a position concerning their eventual subscription to the SDDS. A summary of the TCAP would normally be conveyed to the Board for information as part of background documentation relating to use of Fund resources or Article IV consultation papers.




Developments affecting the Standard

57. Several recent initiatives make use of the Standard (see below). The work on standards and codes, including the IMF experimental Reports on Observance of Standards and Codes (ROSCs) is one example of such initiatives. The Basle Committee on Banking Supervision's draft of a New Capital Adequacy Framework and the IMF's new Contingent Credit Line (CCL) are examples of such initiatives where the Standard is being set up to serve a benchmark role.



58. Stage Three of the EMU, which began in January 1999, resulted in new definitions of statistical aggregates and created a substantial break in data for all Euro Area countries. This new arrangement has a direct impact on the SDDS for these countries (see below).




Uses of the Standard

59. A common theme of much of the discussion on strengthening the architecture of the international financial system has been the need to develop and implement internationally recognized standards and codes of good practice. This focus is reflected in work that has been underway in the Fund to develop and strengthen standards in a number of areas, (including the SDDS) in the Board’s decision to monitor observance of the SDDS, and in the establishment of the joint Bank-Fund Financial Sector Assessment Program (FSAP).



60. The international community has also called upon the IMF to prepare and disseminate reports on the extent to which countries observe such standards. In response, the IMF—in collaboration with the countries concerned—has launched an experimental program to prepare ROSCs. Statistical ROSC modules comparing dissemination practices against the SDDS have been prepared for Argentina, United Kingdom, Hong Kong SAR, Czech Republic, Tunisia, and the Russian Federation (the latter not yet completed).



61. The Fund’s own CCLs have, as one of the criteria for eligibility, that the member would have subscribed to the SDDS and be judged to be making satisfactory progress toward meeting its requirements. Similarly, the Basle Committee’s discussion paper on revising the 1988 Capital Accord proposes that subscription to the SDDS be a requirement for a sovereign borrower to be eligible for a risk weighting below 100 percent and that no bank could have a risk weight less than that applying to its sovereign. Staff has indicated to the Basle Committee that assessments of compliance with the SDDS (and other standards) are not ready for use as a basis for capital adequacy requirements. In part, this reflects unresolved issues relating to the nature and use of assessments, but also the fact that the SDDS has not been designed to arrive at a pass/fail quantitative assessment that could be inserted into a capital adequacy calculation.




European Monetary Union: Impact on the SDDS

62. Ten subscribers to the SDDS participate in the EMU.¹⁶ The January 1999 introduction of the euro and the conduct of euro-area-wide monetary policy by the ECB significantly affected these countries’ status with respect to the SDDS.



63. As of January 1999, the exchange rate of the euro, as well as policy variable interest rates, are determined by the ECB, which is also responsible for the dissemination of these data. Also, subscribers no longer calculate national monetary aggregates but provide monthly data to the ECB, which, in turn, calculates euro-area monetary aggregates. As a consequence, substantial revisions to the metadata for each EMU participant describing the changes to data dissemination practices in these areas, as well as the dissemination of the analytical accounts of the national central banks, were required. The Statistics Committee of the ECB collaborated with the IMF to help subscribers meet their obligations under the SDDS. As a result, revised metadata are being posted on the DSBB. Moreover, the ECB agreed to provide metadata for the euro-area monetary aggregates in SDDS format that could be linked to the DSBB. These additional metadata will provide useful information that complements the information on the metadata base pages of EMU participants.



64. At the request of Eurostat, STA began a collaborative effort to prepare metadata for European Union (the fifteen) and euro area (the eleven) indicators. These metadata are to be posted on a Eurostat web site with links from a euro page on the DSBB to these metadata. These metadata would complement the information already provided by individual SDDS subscribers on their data dissemination practices for national indicators, such as those for output and prices.




III. The General Data Dissemination System¹⁷



A. Report on the Wind-up of Phase I

65. When the Executive Board endorsed the establishment of the GDDS on December 19, 1997, it supported a phased approach to the implementation of the GDDS that focused first on education and training. Phase I, lasting for FY1999 and FY2000, consisted of seminars and pilot metadata development.¹⁸ A series of nine regional seminars was held from July 1998 through November 1999 (see Box 6) to inform countries about the GDDS and to assist them in deciding whether to participate in the System. The Fund invited all non-SDDS subscribers to attend GDDS regional seminars. Country authorities nominated senior officials from two or three of the agencies responsible for production and dissemination of the data covered by the GDDS, such as the national statistical offices, the central banks, and the ministries of finance.
    
    In addition to the seminars, one or two countries in each region volunteered to participate in the preparation of pilot descriptions of national statistical practices (or metadata), including plans for improvement, within the framework of the GDDS (see Box 6). The aim of the regional pilot country work was to offer other countries in the region concrete examples of the type of information expected. National statisticians prepared the pilot metadata in collaboration with staff of the Fund and the World Bank. So far, pilot country work has been completed in nine countries, and pilot metadata, including plans for improvement, are being circulated selectively to other non-SDDS-subscribing countries in the respective region. Work in the remaining four countries is still in progress and is expected to be completed before the end of FY2000.




 

Box 6. Development of the GDDS – Phase I: Training and Information Seminars July 1998: Singapore, for mainland Asia October 1998: Yaoundé, Cameroon, for French-speaking Africa (organized in collaboration with the Banque des Etats de l’Afrique Centrale–BEAC) January 1999: St. Kitts, West Indies, for English-speaking Western Hemisphere (organized in collaboration with the Eastern Caribbean Central Bank–ECCB) January-February 1999: Abu Dhabi, for member countries of the Arab Monetary Fund—AMF (organized in collaboration with the AMF) May 1999: Mexico City, for Spanish-speaking Western Hemisphere (organized in collaboration with the Centro de Estudios Monetarios Latinoamericanos—CEMLA) July 1999: Vienna, Austria, for member countries of the Commonwealth of Independent States—CIS July 1999: Vienna, Austria, for Central and Southern Europe October 1999: Singapore, for Pacific islands November 1999: Pretoria, South Africa, for English-speaking Africa (organized in collaboration with the South African Reserve Bank) Development of Pilot Metadata in Countries Mainland Asia: Bangladesh1 French-speaking Africa: Cameroon2 and Côte d’Ivoire English-speaking Western Hemisphere: Barbados and St. Lucia Member country of the AMF: Kuwait Spanish-speaking Western Hemisphere: Panamá CIS: Kazakhstan Central and Southern Europe: Albania and Bulgaria Pacific islands: Fiji English-speaking Africa: The Gambia and Uganda ___________________________________________________________ 1 GDDS pilot metadata on the fiscal sector are yet to be completed. 2 GDDS pilot metadata related to the fiscal and financial sectors, as well as the socio-demographic area, are to be completed, in collaboration with IMF and World Bank staff, by April 2000.

 

66. Following the Board’s recommendation, close cooperation with regional and other international organizations was actively sought in Phase I. Special mention should be made of collaboration with the World Bank, whose staff conducted the socio-demographic segments of GDDS seminars and assisted in the development of the pilot metadata.



67. The staff prepared the second module of the Fund's Guide to the Data Dissemination Standards. (The first module refers to the SDDS.) The second module, entitled The General Data Dissemination System—referred to hereafter as the GDDS Guide—contains detailed information on the preparation of metadata and other materials associated with GDDS participation. The GDDS Guide was used as a reference in GDDS regional seminars and pilot country work. It is available in English, Arabic, French, Spanish, and Russian. As the GDDS Guide is intended to serve as a reference for a system that is still evolving, it can be expected that it too will undergo change over time.



68. Overall, the impact of Phase I of the GDDS has been significant. In total, 340 senior officials from 129 countries attended regional seminars and 55 countries have notified the Fund that they have nominated a GDDS coordinator to act as the main interlocutor with Fund staff on the GDDS. Of these 55 countries, 13 agreed to be pilot countries and an additional 21 countries have formally advised the Fund that they want to participate in the GDDS. Therefore, a total of 34 countries have expressed an intention to participate in the GDDS, so far.




B. Outline of a Work Program for Phase II of the GDDS

Integration into the work program of STA

69. The implementation of Phase II of the GDDS will take the lessons learned during Phase I and the resource constraints faced both by national compilers and by STA into account. Countries wishing to participate in the GDDS should find the examples of pilot metadata helpful in drafting their own metadata with varying degrees of assistance of Fund staff or experts. STA also intends to revise the GDDS Guide to reflect the experience gained so far. Nonetheless, experience with pilot countries suggests that many potential participants are likely to need assistance to draft metadata and to develop comprehensive plans of action to improve their systems.



70. GDDS activities will gradually be integrated into other mission activities, as well as newer activities, in particular, those related to the production of the ROSCs and statistical capacity building in support of the poverty reduction strategy paper (PSRP) process (see below). Benefits derived from this approach are not limited to containing resource costs; the GDDS framework already has proven useful to technical assistance missions, both in assessing the current status of statistical systems and in developing plans for improvement. Mission activity, other than specific GDDS missions, also will be used to help maintain the currentness of the metadata and plans for improvement of participating member countries.




Development of the DSBB

71. Experience with pilot countries and regional seminars, as well as the findings of the consultants study on the enhancement of the DSBB (see below), all point to a strong interest in the dissemination of both metadata and the data to which they relate. From the perspective of data producers, metadata serve to inform the general public of a country’s participation in the GDDS and its willingness to improve its statistical system. Metadata also help users assess how the data meet their needs. Further, feedback provided by technical assistance agencies, including at the Joint UN/OECD/World Bank/IMF Senior Expert Meeting on Statistical Capacity Building held in November 1999, confirms that GDDS metadata, including plans for improvement and technical assistance needs, would be very useful for these agencies to prioritize and coordinate their activities.



72. The staff has initiated work to present GDDS metadata on the DSBB, making sure that users will be able to distinguish between the SDDS and GDDS. It is expected that GDDS metadata will be posted on the DSBB later this year, as soon as a critical mass of about 10 participating countries has approved its metadata for dissemination on the DSBB.



73. It is expected that the presentation of the GDDS on the DSBB will be enhanced in line with improvements made in the SDDS presentation on the DSBB, as appropriate. However, at this time, there are no plans to link actual data with the corresponding GDDS metadata through hyperlinks to national web sites. The technological constraints on countries that are likely to participate in the GDDS are considerable, but the issue will be reviewed as more experience is gained.




Future development of the GDDS

74. The GDDS is primarily focused on improvements in the production of reliable data, with dissemination issues, especially with regard to the timeliness and periodicity of data, as important but lesser considerations. Experience during Phase I of the GDDS suggests that there is broad support for this strategy, as well as for the coverage and level of detail provided for under the GDDS. The consensus along these lines is facilitated by the fact that the GDDS is less prescriptive than the SDDS and that it is open-ended concerning when objectives are to be achieved. It is not proposed to change the GDDS in this regard, or add to its complexity, which would be too ambitious for many of the countries for which the GDDS was conceived. Member countries may, however, choose the GDDS as a framework for improving their statistical systems as part of any structural adjustment program.



75. Nonetheless, because the GDDS should reflect changing needs for statistics and because, in some cases, it could serve as a stepping stone for countries to subscribe to the SDDS, recent and proposed changes in the SDDS need to be addressed.



76. The data template on international reserves is a newly prescribed category under the SDDS. It may not be advisable to impose a considerable additional dissemination burden on GDDS participants until more lessons can be drawn from experience under the SDDS. Therefore, the staff does not propose to make changes at this time to the data category for international reserves in the GDDS, which includes gross official reserves denominated in U.S. dollars as a core indicator, and reserve-related liabilities as an encouraged category. Nonetheless, should the Fund decide to include the data template among the reporting requirements to the Fund for member countries, the need for building the statistical capacity to compile these would need to be recognized in the GDDS, and the data template included in the encouraged data categories for the external sector.



77. The staff does not propose to strengthen the GDDS with respect to the IIP at this stage, because many member countries are still at an early stage of working toward the objectives included in the GDDS. Instead, the staff proposes to focus on the external debt components within the IIP. Specifically, it is proposed to strengthen the data category for external debt in the GDDS in a graduated manner, consistent with both the needs and strengths of likely GDDS participants. The proposal is based on several considerations. First, these countries tend to have a relatively high concentration of external liabilities in the form of debt instruments—and tend to hold less external assets. Second, many countries that are likely to participate in the GDDS already have invested in external debt monitoring systems, often with technical assistance from abroad, and, therefore, may be able to comply more readily than some industrial countries, especially for public and publicly guaranteed debt. The findings of the external debt survey conducted among SDDS subscribers are consistent with this conclusion (see above).¹⁹ Finally, countries’ capacity to improve external debt data should benefit from the Debt Guide and ongoing seminars and technical assistance.



78. Taking into account the above, the staff proposes to upgrade specific components of external debt from an encouraged to a core data category of the GDDS. The proposal is to include in the core data category public and publicly guaranteed external debt and the associated debt service schedule. The stock data, broken down by maturity, should be disseminated with quarterly periodicity, and timeliness of one to two quarters after the reference date. The associated debt service schedule should be disseminated twice yearly, within three to six months of the reference period, and with data for four quarters and two semesters ahead. Recognizing that non-guaranteed private debt is much more difficult to estimate and typically much less important than for SDDS subscribers, it is recommended that non-guaranteed private debt and debt servicing schedules remain encouraged data categories, with annual periodicity, and disseminated within six to nine months after the reference period.




C. Relationship with Other Initiatives and Organizations

79. The GDDS has been associated with other international initiatives, most notably the development of ROSCs, and Fund staff plans to continue collaboration with a number of other international agencies on the development of GDDS metadata in the coming year.



80. Experimental ROSC modules for statistics based on the GDDS have been prepared and disseminated for three countries, viz. Albania, Bulgaria and Uganda, and two more are planned at this time.²⁰ The experience has shown the usefulness of the GDDS as a framework for comparing country statistical practices with broadly accepted good international practice.



81. The Fund is participating in the work of the Partnership In Statistics for Development in the 21^st Century (PARIS21) Consortium, which was created at the November 1999 Joint UN/OECD/World Bank/IMF Senior Expert Meeting on Statistical Capacity Building. The objective of the Consortium, in which international as well as bilateral donor agencies and recipient countries will participate, is to promote and coordinate statistical capacity building programs in countries preparing PSRPs and similar programs. For this purpose, the Consortium will provide guidance on the data requirements of policy frameworks, the building of strategic statistical master plans, and effective technical cooperation modes. Also, lessons learned from GDDS mission activities are expected to provide useful inputs for PSRPs and the PARIS21 process.



82. As noted above, several international and regional organizations, including AFRISTAT, the Asian Development Bank, the Central American Monetary Council, the Eastern Caribbean Central Bank, the Southern African Development Community, the United Nations Statistics Division, the United Nations Economic Commission for Africa, the United Nations Economic and Social Commission for Asia and the Pacific, and the World Bank have shown a continued interest in working with the Fund on the GDDS. The advantages of collaboration are substantial, as these agencies can bring to bear relevant regional and functional expertise not available at the Fund—for socio-demographic data—on the development of the GDDS. The extent of their involvement, however, may be affected by budgetary constraints, both for them and the Fund.




IV. Trends Affecting Future Developments

A. The Dissemination Standards Bulletin Board

Enhancing the DSBB

Introduction

83. Given the prominence of the DSBB in the implementation of the SDDS and the rapidly growing importance of the Internet, especially the World Wide Web, in the dissemination of economic and financial information, the staff believes that it is imperative that the DSBB becomes a state-of-the-art, user-friendly web site. Only such a site could ensure continued growth in user access to the DSBB. Therefore, an interdepartmental working group was formed to advise on the terms of reference for a study aimed at enhancing the DSBB, and, subsequently, the Patricia Seybold Group (PSG) was hired to conduct an in-depth study of the site and make recommendations for improvements.



84. PSG submitted its final report at the end of January 2000. The study took as its starting point the needs of users of the DSBB and, to this end, conducted market research—Internet surveys and focus groups—to identify these needs. The research revealed the following preferences of DSBB users:



• Access to actual time series data



• Links between the data and the metadata



• Notification of data revisions, release of new data, and noncompliance with release dates



• Access to tools to analyze data and generate reports



• Enhanced capabilities for searching and querying the contents of the DSBB



• A public discussion forum on the site and automatic e-mail links to data providers.



85. In light of these preferences, PSG developed a set of technical and marketing recommendations aimed at strengthening the presence of the DSBB on the Web. While most of the recommendations address technical and noncontroversial issues, some point to possible future scenarios for the development of the DSBB and, therefore, the staff is seeking the Executive Board’s guidance in these areas. Below is a very brief review of the recommendations, including those on which guidance is sought.




Recommendations

86. When introduced in 1996, the DSBB was typical of most first-generation web sites, which featured links to static pages. However, over the ensuing years, the underlying technology of the Web has changed dramatically, allowing a much greater degree of interaction between users and web sites. The PSG made a number of recommendations for enhancements to the DSBB to leverage the Fund’s and its members' initial investment in the development of the site and the underlying data and metadata. These recommendations include (1) improving the interactive and database capabilities of the site so that users could conduct structured queries on the underlying information and receive updates when new information is posted; (2) introducing state-of-the-art technology for the transmission of information between subscribers and the Fund so that the information on the DSBB remains as current as possible; and (3) conducting an aggressive marketing campaign to raise the profile of the DSBB on the Web. All of these recommendations are being given careful consideration by the staff and improvements to the DSBB, based on these recommendations, will be introduced in due course (budgetary resources permitting).



87. Although charging for access to the DSBB is technically possible, it is not clear that there would be sufficient demand to maintain even present use, much less permit greater use in the future, without providing large amounts of actual data as well as metadata on the site (see below). Moreover, users also would expect access to interactive database management tools as well as actual data. While providing such functionality is possible, it raises difficult issues. One example is the question of whether charging for access to the DSBB makes good public policy sense. The public good character of the DSBB in the new information architecture weighs heavily against charging for access to its contents. The staff recommends that the IMF not transform the DSBB into a fee-for-access environment, because such an approach would appear inconsistent with the Fund’s broad objective of ensuring transparency in the dissemination of financial and economic statistics.



88. In view of users' demands, the ability to raise the profile of the DSBB on the Web in the current setting will be closely tied to the degree to which it is associated with actual data. Based on the recommendations of the PSG report, there are three broad approaches that could be adopted to guide the development of the DSBB in the future and on which the staff seeks the Board’s views.



89. The first approach would be to maintain the status quo, i.e., only work toward the completion of NSDPs and hyperlinks from the DSBB to these pages. While this approach ensures a continuous association of the DSBB with actual data, it runs the risk that users will seek more friendly and interactive venues for accessing data, i.e., venues that permit easy downloading into a database of time series data for multiple countries for cross-country comparisons and statistical analysis. In the process, less attention might be paid to metadata. While the status quo is the approach originally envisaged at the time that NSDPs were launched, developments on the Web have facilitated easier and more user-friendly access to multiple data series in the ensuing years.



90. A second approach would be to redisseminate NSDP data on the DSBB, i.e., create a database of the data available on NSDPs at any given time.²¹ The database would be regularly updated—say daily—and allow users the convenience of accessing only one source to view, download, and analyze all the data series available on the NSDPs linked to the DSBB. This approach would strengthen the association of the DSBB with actual data. However, since the database would be redisseminated on the DSBB, it would increase the risk of users confusing these national data with data maintained by the IMF such as International Financial Statistics (IFS) and Recent Economic Developments (REDs). Also, if this approach proved popular, it could lead to user demands for access to longer time series, which would raise serious database management issues and could be opposed by subscribers already providing long-term time series on their web sites, sometimes for a fee. Finally, while the technology exists to allow automatic searching of the various NSDPs to retrieve these data, resource-intensive processes would still be required to ensure the accuracy and reliability of the retrieval process.



91. A third approach would be to reach agreements with commercial vendors that would provide for hyperlinks from the DSBB to their web sites on condition that the vendors agreed to disseminate the SDDS data categories. Vendors would maintain on-line data warehouses (databases) of SDDS data. Vendors would enter into their own agreements with subscribers to secure access to the data, which could be accessed from the DSBB. The overall impact would be to maintain a strong association between the DSBB metadata and the actual data disseminated. The Fund could ensure free access to the data disseminated by subscribers on their NSDPs by continuing to maintain hyperlinks to these pages, while commercial vendors could provide access to longer and more detailed time series data. On the downside, there could be a negative perception of the Fund arising from the association of the DSBB with commercial interests. However, continued free access to the DSBB, as well as hyperlinks to the NSDPs, should maintain the public good character of the SDDS.



92. The staff believes that to ensure the successful future development of the DSBB it is imperative that its association with actual data be further strengthened. Therefore, work to complete hyperlinks to NSDPs for all subscribers will be completed as soon as possible and, to facilitate this process those subscribers who have not established a draft NSDP should do so quickly. Nevertheless, in today’s Internet environment, such links are not likely to be sufficient to maintain a strong interest in the DSBB over time. The costs to the Fund of maintaining and redisseminating NSDP data via the DSBB are likely to be considerable. Moreover, such efforts, if successful, are likely to lead to demands for even more data that could push the Fund into competition with commercial vendors. Given these considerations, the staff believes that the best long-term strategy would be to seek agreements with commercial data vendors that would provide for hyperlinks from the DSBB to their web sites on condition that the vendors agreed to disseminate SDDS data.




B. Data Quality

93. The underlying approach in the Fund’s data dissemination initiatives is consistent with the multidimensional concept of data quality around which a consensus is emerging. By providing users with key information on data characteristics, privileged access and integrity/objectivity aspects, in addition to the summary methodology statements, including statistical techniques, and accuracy cross-checks, the SDDS and the GDDS allow users to be better equipped to make their own assessments of the quality of the data. Moreover, by providing encouragement to and serving as a framework for the improvement of data and dissemination practices, the SDDS and the GDDS also serve as an encouragement for data producers to improve their statistical systems and products.



94. Feedback received on the Fund data dissemination initiatives clearly indicates that users view data quality as important and, therefore in addition to wanting easy access to data, they select data providers that are credible and provide data they can rely upon. Questions have been raised as to whether the approach followed by the Fund to help users make their own quality judgements could be strengthened. In particular, users asked whether the Fund could do more to assess (validate) the quality of the data posted on the subscribers’ NSDP.



95. In this context, STA has moved forward with its work on data quality.



• First, it is extending earlier work undertaken on a framework for assessing the quality of data used by the Fund in a sample of countries. The aim is to design an integrated and flexible framework under which the quality of data can be assessed more evenly across member countries, including by a broad range of users of macroeconomic data. The framework is to be accompanied by a questionnaire that can also serve as a tool to help data producers, especially from countries with less developed statistical systems, to evaluate their own practices. STA is consulting with outside statistical experts in developing such a framework; as work proceeds, STA expects to undertake an outreach program to solicit feedback from a range of data producers and users.



• Second, STA established a Data Quality Reference Site on the DSBB. The main objective of this site is to foster a common understanding of data quality drawing on contributions from the statistical community. This site is expected to evolve over time, reflecting in part comments and suggestions received.



96. In parallel, STA is experimenting with the application of aspects of the data quality approach to the ROSC’s. In addition to assessing dissemination practices—i.e., the extent to which countries adhere to disclosure standards—STA applies aspects of the data quality approach being developed to assess the quality of the country data used by the Fund. The first example of this approach was in the ROSC for the Russian Federation.




V. Resource implications



A. The SDDS

97. Work is currently underway to strengthen and augment the data standards initiatives in relation to the data template on international reserves and foreign currency liquidity. The preparation of data dissemination modules for ROSCs requires additional resources that already have been noted in STA's Medium-Term Resource Plans for FY2001–2003.²² The work on the external debt category, the macroprudential indicators, and data quality issues might also require additional resources that would need to be addressed at a later stage.



98. Current resources allocated to the SDDS need to be increased, especially if development work recommended by the PSG is to be initiated in the short-term; two staff-years would be required for FY2001–2003. One staff-year is required each for the work associated with the marketing of the DSBB and for work on the enhancement of the capabilities of the DSBB dealing with metadata management tools.




B. The GDDS

99. It is expected that the very strong interest among the membership in improving statistics in the framework of the GDDS will lead to a call for acceleration of work with members on GDDS participation and the associated delivery of TA. Careful integration of the GDDS into other activities of STA, collaboration with other regional and international organizations, and, hopefully, continued support under the Administered Account for Selected Fund Activities—Japan are expected to help STA to service the needs of countries that may want to participate.



100. The proposal for the medium term is to increase the resources devoted to the GDDS by three staff-years in FY2001 and another two staff-years in FY2002, and again in FY2003. This would allow STA to assist about 16 new participants per year to prepare GDDS metadata and plans for improvement and subsequently to assure that the metadata remain current. In addition, some countries may be able to prepare metadata and plans for improvement with little assistance from the Fund. The combined scenario—assuming no further increases in Fund staff past FY2003—would permit most countries that are likely to pursue participation in the GDDS to be brought within the system in the next five years.



101. Member countries have indicated that participation in the GDDS may generate additional demand for TA to improve their statistical systems. To the extent that these demands are not met by other TA providers, the Fund may be called upon to help. However, no specific provisions have been made to increase STA's TA program in the medium-term for this purpose. This may add to the existing imbalance between the demand and supply of TA by the Fund. Also, no specific budget request is contemplated for developing the GDDS presentation on the DSBB, because it is expected that additional expenditures for the SDDS also will cover the needs of the GDDS in this respect.




VI. Issues for discussion

102. The Executive Board’s guidance and views are sought by the staff on the following issues:




A. Evolution of the SDDS

Concerning observance and monitoring of the SDDS

103. Although substantial progress has been made by subscribers in meeting the requirements of the SDDS since the second review, owing to a broad range of constraints—including the diversion of resources to address Y2K conversion and institutional rigidities that hampered coordination among statistics-producing agencies—a large number of subscribers are not yet in observance of the Standard. Moreover, it has not been possible to establish hyperlinks from the DSBB to NSDPs for many subscribers. Nevertheless, the staff believes that structured monitoring of the Standard should begin soon to maintain the credibility of the SDDS and strengthen transparency by alerting users to known instances of nonobservance.



• In these circumstances, does the Executive Board agree that such monitoring should begin at end-June 2000, notwithstanding that hyperlinks to NSDPs for all subscribers may not have been established?



104. In establishing the SDDS, the Executive Board called for reviews at appropriate intervals. There have been three reviews, with this latest review encompassing both the SDDS and the GDDS. In the future, the staff proposes to send to the Executive Board, as the need arises, papers proposing actions for addressing cases of egregious nonobservance.



• In addition to these reports, does the Executive Board agree that the next overall review of the Fund’s data standards initiatives should take place by mid-2001?




Concerning enhancements to the DSBB:

105. The staff believes that it is critically important that the association of the DSBB with actual data be further strengthened in the future if the DSBB is to maintain a significant presence on the World Wide Web. Nevertheless, given the costs of maintaining and redisseminating a database of NSDP data on the DSBB and the likelihood that such action could put the Fund into competition with commercial vendors, the staff does not recommend this approach.



• Therefore, does the Executive Board support the staff’s proposal to explore the possibility of the Fund reaching agreements with commercial data vendors that would provide hyperlinks from the DSBB to these sites on condition that vendors maintain on-line data warehouses of SDDS data? If the Executive Board agrees, the staff would report back to the Board on the status of negotiations, seeking Executive Board approval as necessary.




Concerning reserves template data

106. The staff proposes that the sample form shown in Appendix I of this paper be the standard format for the dissemination of template data by SDDS-subscribing countries and also that the sample form be used for the reporting of template data to the Fund. The reported data will be maintained by the Fund in this common format.



• Do Directors agree with the staff’s proposal?



107. The staff also proposes that the data transmitted to the Fund by SDDS subscribers be re-disseminated over the Fund's external web site in the format of the sample form.



• Do Directors agree that the Fund should redisseminate the template data over the Fund's external web site?



108. At the time of the Executive Board meeting of March 23, 1999, Directors agreed that the prescriptions for timeliness of international reserves data dissemination should be reassessed in the context of the third review of the SDDS.



• In view of the experience thus far in the dissemination of template data, the time needed to develop new systems for the compilation of template data, and taking account of the costs and benefits of increased frequency and timeliness, the staff proposes that the existing prescriptions for timeliness and periodicity not be modified at this time. Do Directors agree with the staff's proposal?




Concerning the external debt data category

109. In presenting its proposals on external debt, the staff has tried to make an appropriate balancing of the need for more timely data by policymakers and market participants against the concerns of some data compilers about national statistical priorities and resource costs, including burdens of compliance. The staff has proposed three elements in defining the specifications for the new SDDS data category, only one of which is a prescribed item.



• In view of the difficulties that some countries have expressed in compiling high-frequency data on external debt, such as ensuring adequate coverage of the private sector, do Directors feel that countries will be able to compile data of acceptable quality with the quarterly periodicity and timeliness that is being proposed?



• Thus, on balance, do Directors support the proposal for:




quarterly data to be disseminated within one quarter of the reference period for the external debt of the general government, the monetary authorities, the banking sector, and other sectors, including by maturity—short- and long-term—on an original maturity basis and by instrument as set out in the BPM5;





dissemination of supplementary information on prospective debt service payments, in which the principal and interest components are separately identified, twice yearly with data for the first four quarters and the following two semesters ahead encouraged;





dissemination of a domestic/foreign currency breakdown of external debt each quarter by sector within one quarter of the reference period encouraged; and





a proposed transition period for implementing these proposals that would run through March 31, 2003?





• Provided that quarterly external debt are disseminated with a one-quarter lag, do Directors support the proposal to change from six months to nine months the prescribed timeliness specification with which IIP statistics are to be disseminated under the SDDS?




B. Evolution of the GDDS

110. The staff plans to continue working with potential GDDS participants to assist them with the drafting of metadata and to develop comprehensive plans of action to improve their statistical systems. In doing so, STA intends to integrate the GDDS closely with other departmental activities in order to maximize efficiencies, reduce costs for the Fund and member countries, and increase the effectiveness of member countries’ efforts to improve their systems and related technical assistance activities. Further, the staff is developing the DSBB web site to include country presentations of GDDS metadata, including plans for improvement. As was requested by Directors at the Executive Board meeting of March 19, 1997, a clear distinction will be maintained between SDDS and GDDS country presentations.²³



• Do Executive Directors agree with the proposal to include GDDS metadata on the DSBB, without hyperlinks to actual statistics, but to adopt other technical enhancements of the web site proposed for the SDDS?



111. The staff proposes to strengthen the external debt data category of the GDDS, but to leave other data categories unchanged at this time.



• Do Executive Directors agree to maintain the distinctions between the coverage of the SDDS and GDDS, and to strengthen the external debt data category of the GDDS, as proposed?