Executive Summary1. This paper responds to requests and guidance provided during Executive Board discussions in December 1997 and September 1998. The Board has stressed, and events in the international financial system have repeatedly confirmed, that provision of comprehensive, timely, and accurate economic data by member countries is essential for effective Fund surveillance. In recognition of this, and in line with members' obligations under the Articles of Agreement, the Board has previously decided that all members should provide regular information on a set of core indicators; and that, on a case-by-case basis, such information should be supplemented by reporting additional relevant data for surveillance, particularly data critical for the early identification of external and financial sector vulnerability. Directors have also endorsed the concept of a minimum standard for the provision of data on reserves and related items, and have requested the staff to develop a template outlining the expected coverage. 2. A review of members' provision of data to the Fund shows that a majority of the membership continue to provide timely information on all core statistical indicators. However, countries that have had difficulties providing such data in the past generally continue to do so. The persistence of these difficulties is primarily attributed to severe resource constraints and long gestation periods required for statistical capacity building. 3. A review of the treatment of data issues in Article IV consultations reveal that the staff and the Board have devoted considerable attention to such issues. Most staff reports contain substantial information on data issues; many of them identify data deficiencies that have limited the effectiveness of surveillance and discuss specific recommendations to address such deficiencies. However, data issues still were not addressed in the summings up of most consultations. In addition, the review pointed to little direct linkage between the treatment of data issues and policy analysis in reports. 4. While reporting on core indicators has become an element of every Article IV consultation report, the characteristics of the data to be provided (such as coverage, frequency, timeliness) have not been defined by the Board. This reflects a shared view that data requirements for surveillance change over time and vary across countries. However, in the wake of recent crises in emerging-market countries, improving information on international reserves and external debt has become a priority in efforts to strengthen surveillance. For these reasons, and following previous guidance from the Board, this paper proposes the establishment of benchmarks for the provision of data to the Fund on reserves and related items and on external debt. In view of the diversity of circumstances facing individual members, the benchmarks would serve as points of reference rather than absolute standards. Staff reports would compare countries' practices with these benchmarks, indicating the reasons for any differences, their significance, and (if appropriate) the member's plans for strengthening data provision in these areas. 5. As regards reserves and related items, the paper suggests adoption of a benchmark that parallels the SDDS prescription—use of the data template on reserves and foreign currency liquidity and related sample form, with monthly periodicity and timeliness. As regards external debt, the paper recommends a benchmark incorporating both the prescribed and encouraged elements of the SDDS data category on external debt. The benchmark would thus entail quarterly reporting of external debt statistics disaggregated by sector, maturity, and instrument; biannual reporting of a debt service schedule for the two years ahead; and quarterly reporting of the domestic-foreign currency breakdown of external debt. 6. The staff considers that treating these detailed specifications regarding reserves and external debt as benchmarks, rather than absolute standards, is critical in view of the diversity of circumstances facing individual Fund members. The constraints faced by members in finding the necessary resources and the time needed to strengthen data systems must be recognized. The staff proposals in both of these areas would imply a need for many, if not most, members to strengthen their data systems. The challenge would be much greater in the case of external debt, particularly as regards debt of the private sector. In addition, while the reserves data template is intended to provoke an examination of risks that might be faced by the authorities in any member country, not all of the information called for under the proposed benchmark for external debt data is of equal relevance to all countries (for instance, the lack of detailed data on private sector debt may not greatly hamper the analysis of external vulnerability in a country where the private sector has little or no access to external borrowing). In other instances, data required for adequate surveillance may need to be more timely or more detailed than specified in the benchmarks. 7. The Fund will likely be requested to offer technical assistance to national authorities in support of their efforts to strengthen compilation and provision of better data on reserves and external debt, and in other important areas. The needed resources could be substantial, particularly as regards external debt where the work is less advanced. In addition to Fund staff resources, support for technical assistance might be available from the other agencies that participate in the work of the Inter-Agency Task Force on Finance Statistics. 8. Substantial deficiencies in fiscal data often affect analysis of fiscal developments. The forthcoming Manual on Government Finance Statistics will provide an improved conceptual framework to help address these deficiencies. However, besides better methodology, resolution of existing difficulties will also require sizeable and sustained efforts to strengthen data systems. This paper asks whether staff should pursue technical work on issues related to the eventual development of a benchmark on provision of fiscal data. 9. The Fund is intensifying its efforts to assist countries to improve data quality. The focus is presently on the development of a framework for assessing data quality, which is being designed to be applicable across all macroeconomic sectors and to be used in a flexible manner. In preparing this framework, the staff will consult with outside experts and gather the views of a broad range of data compilers and users. 10. The Executive Board has already had a preliminary exchange of views on the possibility of broadening the application of Article VIII, Section 5 of the Articles of Agreement to help ensure members provide information the Fund needs for its work. This paper identifies approaches that could be considered in this regard, including the option of a decision establishing that all members are required to provide, at a minimum, all core statistical indicators. It recalls the earlier discussion of the possibility of adopting specific decisions on the data to be provided by individual countries, in the context of use of Fund resources, and notes that, as a complement, a similar approach could be considered in the context of Article IV consultations in particular cases. I. IntroductionA. Background11. Events in the international financial system in recent years have highlighted both the role of Fund surveillance in the prevention and management of crises and the importance of adequate economic data as a basis for effective surveillance. At the first review of data provision to the Fund for surveillance purposes in 1995, Executive Directors agreed on members' provision of a set of core indicators for surveillance and stressed the need for these indicators to be supplemented by the regular reporting, on a case-by-case basis, of additional data. At the 1997 review, some Directors suggested that the definition of core indicators be extended to include additional data that would be critically important for identifying emerging signs of external and financial sector vulnerability at an early stage. In a related discussion in September 1998, Directors expressed general support for the proposal for a minimum standard for data provision to the Fund with respect to reserves and related items. 12. This paper is the fourth in the series that reviews current practices with respect to the provision of data to the Fund, and it makes a number of proposals that respond to the concerns expressed by the Board. This section provides a brief review of data provision to the Fund and the treatment of data issues in the Article IV consultation process. Section II discusses proposals for strengthening the provision of international reserves and external debt data. Section III discusses ongoing work to improve fiscal data, initiatives to develop a framework for assessing data quality, and application of Article VIII, Section 5. Section IV presents conclusions and issues for discussion. B. Data Provision and Treatment of Data Issues113. As background for this review, the staff has surveyed the provision of core statistical data by all Fund members, the provision of data on external reserves and related items by a sample of member countries (mainly emerging markets), and the treatment of data issues in Article IV consultations. Country practices with respect to the specific definition, coverage and timeliness of the core statistical data provided to the Fund continue to vary widely, and improvements in the provision of data to the Fund by individual countries since the last survey conducted in 1997 have been modest. Broadly, the results show that those countries that have typically provided the core statistical data to the Fund in a timely fashion (and these are in the majority) continue to do so, while countries that have had problems in providing timely data to the Fund continue to experience difficulties. In general, the difficulties experienced by this latter group of countries can be traced to the severe constraints on the resources they devote to statistics and to the long gestation period required for statistical capacity building. Within these general results a few points stand out:
14. The staff also reviewed the treatment of data provision to the Fund in summings up and staff reports for all Article IV consultations concluded in the period March 1999-February 2000, covering 131 countries and regions. This review confirmed that considerable attention was devoted by both the Board and the staff to data provision to the Fund and related issues, such as data quality, in Article IV consultations. The treatment of data issues in Article IV consultation documents generally included a description of the availability and quality of macroeconomic statistics, information on progress made in addressing past statistical weaknesses, and recommendations to address remaining shortcomings. 15. Three aspects of the sample results deserve special mention. First, as reported in more detail in Appendix I, an explicit general assessment of the adequacy of data provided to the Fund was contained in about 80 percent of the Article IV staff reports and about 40 percent of the summings up. This contrasts with previous guidance from the Executive Board relating to the need for inclusion of such assessments in staff reports and summings up. Second, in about 40 percent of the staff reports, this overall assessment indicated that data deficiencies had "hampered surveillance," or "limited economic analysis and policy formulation." Third, while data issues received substantial coverage in staff reports, their treatment was rarely integrated in the detailed discussion of macroeconomic developments and policy issues. In particular, reports seldom attempted to draw attention to uncertainties regarding the analysis of past economic developments, projections, or ensuing policy conclusions stemming from noted data deficiencies. There were no cases in which the staff or the Executive Board stated that a member was not cooperating in providing data for surveillance purposes. II. Strengthening Data Provision to the Fund16. The earlier Executive Board discussions on data provision have focused on core indicators for surveillance and have drawn attention to the possibility of augmenting such indicators or more fully articulating existing indicators to increase the effectiveness of surveillance. The lack of current and transparent data during recent financial crises has underscored, in particular, the need for more timely, comparable, and comprehensive information on international reserves and external debt to underpin the assessment of vulnerability. The Board has also urged the staff to develop further its work on the assessment of data quality, with a view to using it to improve the statistical basis for surveillance and to enhance the effectiveness of the Fund's technical assistance. Finally, during its recent discussion of misreporting of information to the Fund, many Directors saw a need to consider updating the set of data members may be required to report to the Fund, pursuant to Article VIII, Section 5. This section of the paper proposes the establishment of benchmarks for the provision of data in the areas of international reserves and external debt and discusses ongoing work to improve fiscal data. A. Benchmarks for Data Provision to the Fund17. At the March 1995 discussion of the provision of statistical data by members and subsequent reviews of the provision of information to the Fund for surveillance (see Box 1), the Executive Board identified and confirmed the core indicators as the minimum set of data that should be provided to the Fund for surveillance. At the same time, however, the Board has recognized that individual countries' data practices, level of statistical development, and data requirements for surveillance purposes may be different and may change over time. As a result, the characteristics of the data to be provided as core indicators, such as coverage, frequency, and timeliness, have not been specifically defined.
18. In view of the importance of data on reserves and external debt for the assessment of external vulnerability, as well as the role of the Fund as a repository of country experience and work on methodological issues, Directors endorsed a broad-based effort to improve statistical information in these areas at the September 1998 discussion of data availability. In that context, Directors expressed general support for the establishment of a minimum standard for data provision to the Fund with respect to reserves and related items, and encouraged the staff to prepare a template outlining the coverage expected. Most Directors indicated that this should be regarded as a benchmark against which to measure performance. While seeking a parallel strengthening of data on external debt, Directors considered that it would be premature to propose an equivalent minimum standard in this area. In the paragraphs below, the staff proposes benchmarks for the provision of data on reserves and related items, and on external debt. These benchmarks would encompass specific provisions relating not only to periodicity and timeliness but also to issues such valuation and coverage, drawing on methodological work undertaken in the context of the Special Data Dissemination Standard (SDDS) and the General Data Dissemination System (GDDS). 19. The staff considers that treating these detailed specifications regarding reserves and external debt as benchmarks, rather than absolute standards, is critical in view of the diversity of circumstances facing individual Fund members. To implement the staff proposals in both of these areas, there may well be a need for most members to strengthen their data systems. The challenge would be much greater in the case of external debt, particularly as regards the coverage of private sector debt. In addition, while the coverage of the reserve template is intended to provoke an examination of risks that might be faced by the authorities in any member country, issues of selectivity in the coverage of debt data may arise as, for instance, obtaining data on private sector debt may simply not be a high priority in countries where the private sector lacks the capacity to borrow significantly abroad. By the same token, the data required for adequate Fund surveillance in some cases may be more detailed and timely than what is implied by the benchmark. In this sense, the benchmarks would be neither a compulsory floor nor a ceiling. Staff reports would compare countries' practices with these benchmarks, indicating the reasons for any differences, their significance, and (if appropriate) the member's plans for strengthening data in this area in the future. This evaluation of provision of reserves and debt data would take fully into account the diversity of circumstances faced by members as highlighted above, including issues of relevance and technical constraints affecting the compilation and production of such data. 20. It should be stressed that the call for strengthening data provision to the Fund in this manner for reserves and external debt was made in the context of an expectation that efforts by members would be supported by a wide-ranging effort to clarify methodological standards and make appropriate technical assistance available. It will be crucial that, over time, the expert resources available collectively to the Fund and other participating institutions be commensurate to this task and, perhaps more importantly, that national statistical agencies in member countries be adequately staffed and equipped. 21. The staff would not propose the specification of comparable benchmarks at this time for other core data for surveillance. While recent developments in Fund surveillance and monitoring of Fund-supported programs have highlighted the need for a significant strengthening of fiscal data, the work on assessing the experience in member countries and on developing appropriate methodological standards and data systems in areas other than reserves and external debt is not sufficiently advanced to warrant the specification of benchmarks for these other core data. This paper provides information on work underway to identify necessary improvements in methodology and data systems in the fiscal area. B. A Benchmark for Reserves22. Following the September 1998 discussion of data availability, dissemination, and provision to the Fund, the staff, jointly with the BIS Committee on the Global Financial System (CGFS), developed a template for the compilation of data on international reserves and foreign currency liquidity. This template was refined in subsequent discussions and eventually adopted by the Executive Board as the part of the prescription in this area under the SDDS. The Fund staff subsequently developed operational guidelines for the compilation of the template data and a sample form for submission of these data to the Fund, in consultation with a wide range of members and other international institutions.2 23. The survey of current practices in reporting data to the Fund on reserves and related items, in the previous section, revealed wide differences in the classification, coverage, frequency, and timeliness of the data and concerns about data quality. While recognizing that some differences are to be expected, given the diversity of the Fund's membership, data on reserves and related items are fundamental tools for the design and implementation of economic policy and the assessment of vulnerability. The use of a common point of reference, such as the template, could bring significant improvements across countries in judging the comprehensiveness, comparability, and quality of these data. 24. Against this background the staff proposes that the SDDS prescriptions regarding data on reserves and foreign currency liquidity-including the data template, the use of the sample form as a format for reporting, and monthly periodicity and monthly timeliness-be adopted as a benchmark for the provision of data on reserves and related items to the Fund for surveillance purposes. Methodological issues would be treated as under the Operational Guidelines for the SDDS data template. 25. The experiences of SDDS-subscribing countries so far with regard to the template data are instructive. Almost all subscribers have had to take steps in preparation for observing this prescription, such as establishing coordination mechanisms to gather data from various sources to meet the institutional coverage specified in the template, as well as developing methods and procedures to reconcile accounting data provided by various agencies with the underlying concepts of the template. A number of countries are still grappling with these issues. The transition period ended on March 31, 2000, implying that all subscribers were expected to disseminate data in this form for end-April, no later than May 31, 2000. A preliminary review of data disseminated according to the template suggests that most SDDS subscribers have provided comprehensive data on their international reserves and related items. The authorities of countries that have not yet disseminated this data cite technical problems or resource constraints as having temporarily delayed the dissemination of the data. They have been in contact with Fund staff to resolve these issues. These experiences and the fundamental diversity of the membership reinforce the view that the SDDS prescriptions regarding data on reserves and foreign currency liquidity should be viewed as a benchmark for comparison that should be applied with due regard for the situation of the member as well as the need for adequate information for surveillance purposes, and a view toward improvement, as warranted, over time. C. A Benchmark for External Debt26. At the 1998 Board discussion, Directors also noted serious limitations to existing data on external debt and identified improvements in the compilation of external debt data as a high priority. Given the limitations of the underlying data systems at that time, Directors were of the view that it was premature to propose a minimum standard for provision of data on external debt to the Fund for surveillance. Since then, intensive efforts have been made by the Fund, in cooperation with other institutions, to foster improvements in the coverage and compilation of external debt statistics. The importance of timely and comprehensive external debt statistics has also been reinforced in various other international initiatives launched since 1998, in particular through the work of the Inter-Agency Task Force on Finance Statistics, the Financial Stability Forum's Working Group on Capital Flows, and the February 2000 Conference on Capital Flow and External Debt Statistics held at the IMF.3 Against the background of these developments, at its meeting on March 29, 2000 for the Third Review of the Fund's Data Standards Initiatives, the Board took a number of decisions to strengthen the requirements of the SDDS and GDDS with respect to external debt statistics (see Box 2). 27. In the area of external debt statistics, the core indicator for data provision to the Fund has never been precisely defined. However, in most cases the coverage has been limited to the stock of public and publicly guaranteed external debt. Because of the growing importance of private sector external debt, this narrow coverage is no longer adequate for conducting surveillance in many countries. Moreover, in some cases, the debt data reported to the Fund have been found to be inadequate even for the public sector—for instance, because external debt in the form of domestically issued debt securities held by nonresidents was not being recorded (for instance, in Russia, prior to the 1998 crisis). 28. Thus, although the core indicators table included in the Article IV consultation report may indicate that a country is reporting timely and frequent external debt data to the Fund, the data may not be comprehensive. The value of the indicator is thereby diminished as is the Fund's ability to conduct effective surveillance. Moreover, as the data are not reported according to any agreed international standard, there are problems of comparability across countries, which further diminishes the analytical value of the data.
29. In view of the critical importance of detailed, comprehensive, and timely external debt data for effective surveillance, the staff proposes that the core indicator for external debt be more fully articulated, based on the SDDS prescription. The staff proposals should be seen as a benchmark with which to compare countries' data provision practices to the Fund, rather than as a specific requirement, for the reasons discussed above. The staff proposes that countries:
30. Work of staff and other agencies shows that it will likely take many years for most countries to provide the full range of external debt data identified as important for surveillance. This view is underpinned by the results of the 1999 Fund survey of data compilers in 47 SDDS-subscribing countries.4 It was found that 70 percent of the developing/emerging market countries that subscribed to the SDDS compiled and disseminated either monthly or quarterly external debt or international investment position statistics. Only 35 percent of the industrial countries disseminated monthly or quarterly data in these areas. Reflecting the magnitude of the improvements in data systems that are required, the transition period for implementing the SDDS prescription on external debt runs through March 31, 2003. 31. The responses to the survey also revealed that only 40 percent of the SDDS subscribers were in a position to compile elements of an amortization schedule. Frequently, information was not available for the full range of debt components. For example, data on forward amortization payments were frequently not available for trade credits, deposits, and non-equity liabilities to foreign direct investors. It must be recognized that the development of debt service schedules is likely to be an especially difficult task in countries that do not collect basic debt information at the individual loan/security level. More broadly, this situation reflects the various circumstances, national statistical priorities, and constraints that compilers face, as well as the inherent difficulties of compiling external debt statistics, such as achieving complete coverage of the obligations of the private sector, measuring short-term trade credit, determining the residency of holders of domestically issued debt securities, and recording complex financial transactions such as repurchase agreements. 32. The proposed benchmark is demanding, in that it includes both the prescribed and recommended elements for the dissemination of external debt data under the SDDS. The benchmark is pitched at information that could make a crucial difference in identifying external vulnerability in most member countries, based on the experience with recent financial crises. In interpreting countries' practices in relation to the benchmark, the staff and Executive Board will need to bear in mind that the staff proposal represents a level of statistical development in this area that will be difficult for many countries to achieve. In addition, as noted above, certain specific components proposed in the benchmark may not be relevant for some countries. For other countries—for instance, those experiencing or recently emerged from financial crises—there may need to be a major, high-priority effort to bring the external debt data up to the benchmark and to develop higher-frequency reporting on some key items. Such considerations would need to be taken into account in assessments of the provision of external debt data for surveillance. Finally, the need to provide adequate resources for compilers to produce the data required for surveillance has to be recognized by policymakers. D. Resource Implications33. Experience with SDDS subscribers suggests that the Fund may need to offer technical assistance to national authorities to familiarize them with the concepts and coverage of the data template for reserves and international liquidity, to assist them in resolving specific methodological issues, and to explore possible data sources. In the area of external debt, work is less advanced. Thus, the resources and the time frame required to assist countries to move toward that benchmark is likely to be more substantial. 34. If work is to proceed quickly to assist country authorities in their efforts to improve data provision in the area of reserves and external debt—a highly specialized area—significantly more resources will be needed. If additional resources are not made available, but priority is to be given to this area, such work would inevitably come at the expense of technical assistance in other areas. Support for technical assistance may be available from the other agencies that participate in the work of the Inter-Agency Task Force on Finance Statistics, such as the BIS, the Commonwealth Secretariat, the OECD, UNCTAD, and the World Bank. 35. Staff's proposals to establish benchmarks for the provision of reserve and debt data are designed to lead to data provision to the Fund that is more comprehensive, timely, and accurate. In many instances, additional efforts to achieve this goal will entail greater use of staff resources on data issues in area and functional departments in the context of regular surveillance activities. III. Other Issues Related to Data Provision to the Fund36. This section discusses ongoing work to improve fiscal data, staff initiatives to develop a framework to assess data quality, and application of Article VIII, Section 5.5 A. Efforts to Improve Fiscal Data37. Recent developments in Fund surveillance and implementation of Fund-supported programs have also led to an increased awareness of deficiencies in the reporting of fiscal statistics. 38. Comprehensive fiscal data for the central government and general government sectors are reported by a number of countries participating in the SDDS. Coverage issues in government finance statistics, particularly relating to state and local governments, have, however, been problematic for a number of these countries. Even advanced industrial economies often find it difficult to supply comprehensive fiscal data on a timely basis. For example, although some countries have begun to implement accrual accounting, they have represented the SDDS specifications for timeliness and/or periodicity as unsuitable for the greater complexity involved in accrual accounts. In many countries, the pursuit of quasi-fiscal and regulatory activities through nonfinancial public enterprises and public financial institutions compounds the difficulties of obtaining an adequate measure of the fiscal balance for analytical purposes 39. For non-SDDS subscribers, the fiscal data provided to the Fund often are also subject to qualitative weaknesses.6 These may include an inability to reconcile the "above the line" deficit derived from identifiable revenue and expenditure items with the identifiable "below the line" financing items derived from monetary survey or balance of payments data. Changes in stock data (such as the amount of government debt) may not reconcile with flow data (e.g., on the fiscal deficit). Overlaying these problems are poor accounting practices and differences in the reporting bases (commitments, accruals and cash reporting). All these problems may make it difficult to meaningfully reconcile fiscal data reported to the Fund with the generally more accurate monetary and external account data produced. 40. Several of the deficiencies in fiscal data provision cannot be easily resolved within the current government finance statistics methodological framework which is widely used by most countries. The forthcoming GFS Manual—being developed in close collaboration with statisticians in a number of countries—provides an improved conceptual framework for fiscal data. Fundamental to the improved methodology is the use of accrual—rather than cash-based—accounting. 41. However, it is important to be realistic about both the speed at which the required improvements can be implemented and the extent to which country officials and the Fund staff can overcome the aforementioned problems. Strengthening fiscal data systems will require costly investments by member countries to build the necessary information system infrastructure, supported by training and technical assistance spanning several years. In coming months, the staff will also consider whether the Fund should seek to come up with benchmark best practices on data provision in the fiscal area, given the many inherent difficulties. B. Data Quality42. In recent years the Fund has intensified efforts to assist countries to improve the quality of their data. This has been accomplished through a variety of complementary efforts, including the expansion of the work of the Statistics Department on statistical methodologies, technical assistance and training activities, greater focus on the adequacy of member countries' data for surveillance, and the Fund's data standards initiatives. 43. As reported to the Board on the occasion of the Third Review of the Fund's Data Standards Initiative, the Statistics Department is developing a framework for assessing data quality. The main objective of this framework is to provide more structure for the assessment of the quality of data, including those provided to the Fund for surveillance.7 The framework combines a quality-of-the-system approach with a more explicit quality-of-the-product approach, which emphasizes accuracy and reliability.8 44. Under this framework, the quality of statistical information includes such factors as integrity, accuracy, coherence/consistency, timeliness, understandability, and accessibility. These factors can be decomposed into observable and, to the extent possible, quantifiable indicators and practices against which to assess data quality. The framework is designed to be sufficiently comprehensive to be applicable across macroeconomic sectors, and is meant to be used in a flexible manner. 45. In developing the data quality assessment framework, the Statistics Department is consulting with outside statistical experts and will undertake an extensive outreach program to gather the views of a broad range of data compilers and users. The overall approach, relevant framework and interpretation guidelines for individual data categories are expected to be discussed at a number of forthcoming international meetings on statistical issues. In parallel, the assessment framework is being tested by the Statistics Department's technical assistance missions. C. Application of Article VIII, Section 546. During its recent discussion of misreporting of information to the Fund, the Executive Board had a preliminary exchange of views on application of Article VIII, Section 5, as part of a range of actions designed to help ensure the accuracy of information reported to the Fund.9 Directors stressed that Article VIII, Section 5 should be applied in a manner consistent with its relevance to the modern data needs of the institution. In particular, many Directors saw a need to update the set of data members are required to report to the Fund pursuant to this Article. These Directors cautioned, however, that the Fund's information requirements could not be specified once and for all. They also noted that the modalities for a possible broadening of the application of Article VIII, Section 5 would have to be carefully considered to take into account, inter alia, members' varying degrees of ability to provide data. 47. In order to facilitate further discussion of the application of Article VIII, Section 5, this section of the report presents information on relevant elements of the approach followed since 1995 regarding provision of information to the Fund for surveillance. 48. The categories of information identified in 1995 as the core minimum for effective Fund surveillance extend significantly beyond the types of information specifically listed in Article VIII, Section 5. In particular, the data categories listed in Article VIII, Section 5 (a) do not include reserve or base money, broad money, interest rates, the central bank balance sheet, and the fiscal balance. In addition, Article VIII, Section 5 (a) makes no reference to other data categories that, in some cases, have recently been assessed to be essential for the conduct of effective surveillance, such as financial sector data. 49. While the Executive Board has generally agreed that members have to abide by certain minimum requirements for provision of data to the Fund, Executive Directors have stated that a single standard for all Fund members would be neither desirable nor realistic. In coming to this view, Directors have noted that data requirements change over time. They have also observed that both the types of data that are necessary to monitor and analyze developments and the availability of data vary from country to country. Thus, cross-country variations in the scope and frequency of data to be provided to the Fund may be fully compatible with the Fund's responsibility to conduct surveillance in an evenhanded manner. 50. In order to resolve cases in which inadequate provision of data to the Fund was seen to stem from reluctance on the part of the member, while striking an appropriate balance between strengthening data provision to the Fund and fostering cooperation with members, the Executive Board has endorsed a graduated and cooperative approach. This is to be applied on a case-by-case basis involving, successively, direct staff and management contacts with the authorities, assistance from the Executive Director concerned, consideration of the matter by the Executive Board, and the possibility of a decision to postpone the conclusion of an Article IV consultation. 51. In contrast, the Executive Board has not given consideration to the use of the sanctions for breach of obligation under Article VIII, Section 5. This may be related to the fact that, with one exception in the 1950s, no case of such breach in the context of surveillance has come to light. 52. The staff would suggest that, in discussions of application of Article VIII, Section 5, further consideration be given to the following elements:
53. Consideration could also be given to adoption of specific decisions requiring the provision of information that the Fund regards as being especially important for particular cases, in the context of decisions on use of Fund resources by individual countries. Such decisions could cover the technical features of the reporting on the data categories listed in the uniform decision as well as reporting of data not included in those categories. Adoption of specific decisions of this type in the context of Article IV consultations could also be considered in particular cases. Such an approach would be prospective, involving a decision at the time of an Article IV consultation on the data requirements for subsequent consultations. The staff will return with specific proposals in this area, taking into account Directors' views in the discussion of this paper. IV. Issues for Discussion54. Executive Directors may wish to focus their discussion on the issues summarized below. Data provision and treatment of data issues 55. Staff surveys indicate that the majority of countries have continued to provide the core statistical indicators to the Fund in a timely fashion. However, for some countries, resource constraints and the long time required for statistical capacity building have hampered the timely provision of data to the Fund. 56. Staff reports for Article IV consultations generally devote considerable attention to data provision to the Fund and related issues. However, a majority of summings up still do not make any reference to data issues. Moreover, the treatment of data issues in staff reports is not always integrated with the discussion of macroeconomic developments and policy issues. Do Directors agree that further progress in the provision of core statistical indicators is needed and that countries should be encouraged in their efforts to press forward with statistical capacity building? How do Directors assess the coverage of data issues in Fund surveillance? Do Directors still consider that summings up should include a paragraph on data provision to the Fund in all cases? Do they agree that the treatment of data issues should be more closely integrated in the discussion of macroeconomic analysis and policy issues? The establishment of benchmarks for provision of data on reserves and external debt 57. In light of the importance of comprehensive and timely data on external reserves and related items and on external debt, the staff is proposing the establishment of benchmarks with regard to data provision to the Fund for surveillance purposes in these two areas. As described in paragraph 19, the benchmarks would be neither compulsory floors nor ceilings; they would serve as points of reference rather than absolute standards. Staff reports would compare countries' practices with these benchmarks, indicating the reasons for any differences, their significance, and (if appropriate) the member's plans for strengthening data in this area in the future. Do Directors agree with the proposed general approach to the establishment of benchmarks in these crucial areas? A benchmark for reserves data 58. Directors previously expressed general support for a minimum standard for data provision to the Fund with respect to reserves and related items and encouraged staff to come back to the Board with specific proposals in this area. In view of the diversity of members' situations, the staff proposes a benchmark in this area that is consistent with the corresponding SDDS prescriptions, as described in paragraph 24. Are Directors in agreement with this proposal? A benchmark for external debt data 59. In recent years, intensive efforts have been made to foster improvements in the coverage and compilation of external debt statistics by member countries. In light of these developments, the staff proposes to establish a benchmark for provision of external debt data to the Fund as described in paragraph 29. Are Directors in agreement with this proposal? Resource costs 60. The initiatives to establish benchmarks for the provision of data to the Fund in the area of reserves and external debt will have resource implications that will need to be made explicit in the Fund's budget. Do Directors agree that priority should be given to making available the resources to support technical assistance to aid countries in working toward meeting the benchmarks? Efforts to improve fiscal data 61. The report indicates that both methodological advances and data system strengthening are needed to overcome existing deficiencies in fiscal data provision. Do Directors support continued work by staff on methodological issues related to the eventual development of a benchmark on provision of fiscal data? Data quality 62. The staff is developing a framework for assessing the quality of data used for macroeconomic analysis, in consultation with outside experts, to contribute to more uniform assessments of data quality by staff (Section III.B.). When fully developed, this framework would be expected to provide a basis for the assessment of data systems and series in a flexible manner appropriate to the circumstances of individual countries. At the conclusion of the third review of the Fund's data standard initiatives, Directors endorsed the staff's efforts to develop such a framework. Do Directors still consider this endeavor to be a high priority? Coverage of Article VIII, Section 5 63. The report suggests elements of an approach for application of Article VIII, Section 5, as described in paragraph 52. Do Directors agree that these elements provide a basis for further consideration of this issue? Future reviews of data provision to the Fund 64. This paper is the fourth review of data provision to the Fund. Do Directors agree that such reviews are useful and that the next review should take place in about one year's time to assess, inter alia, the experience with the benchmark approach? Appendix 1DATA PROVISION AND TREATMENT OF DATA ISSUES 65. The current operational framework for the treatment of data provision to the Fund in Article IV consultations was put in place in 1995. In the wake of the Mexican crisis and in the context of discussions on strengthening surveillance, the Executive Board emphasized the need for provision of complete and consistent data by members in the Article IV consultation process. Executive Directors also stressed the importance of a regular flow of key information between Article IV consultations for surveillance to be effective, continuous, and timely. Reflecting the priority attached to data provision, the Executive Board requested that all Article IV consultation reports contain an assessment of the data provided to the Fund. Specifically, the staff was instructed to include an explicit assessment of the periodicity, timeliness, coverage, and quality of the member country's economic data in the introductory section of Article IV consultation reports; identify any concerns about the adequacy of a member's data provision for effective surveillance in the staff appraisal, and include recommendations for improvement; and address such data issues in greater detail in an annex, which would include a matrix summarizing the provision of core statistical indicators. 66. Subsequent reviews of data provision to the Fund by the Executive Board were conducted against the background of the crises in Asia and Russia and the call by Directors for the timely and comprehensive provision of data for effective surveillance and crisis prevention. While reaffirming the broad validity of the 1995 framework, the Board called for its strengthening in specific areas. In the 1997 review, Directors urged staff to continue to provide frank and comprehensive assessments of data issues in Article IV consultation reports. Finally, while noting the modest improvement in the timeliness of the provision of core indicators, Executive Directors thought that they needed to be complemented by other data taking into account the circumstances of individual countries; in that context, Directors emphasized the importance of reserve-related liabilities, central bank derivative transactions, private sector external debt, and prudential-type bank indicators. In September 1998, Executive Directors endorsed efforts to improve the availability and quality of data on international reserves and external debt through a strengthening of statistical systems and methodologies and enhancement of data provision to the Fund in these areas. Directors generally supported the proposal for a minimum standard with respect to data provision to the Fund with respect to reserves and related items and the development of a template outlining the coverage expected. Most Executive Directors also called for the systematic inclusion of a paragraph assessing data provision to the Fund in summings up of Board meetings concluding Article IV consultations. They also generally agreed that, in cases where data provision to the Fund seriously hampered surveillance, the Board should consider its response on a case-by-case basis. 67. In light of the above, the staff has reviewed the provision of core statistical data by all Fund members, the provision of data on external reserves and related items by a sample of emerging-market countries, and the treatment of data issues in Article IV consultations. 68. In providing the core indicators to the Fund, country practices with respect to the specific definition, coverage and timeliness of these indicators vary widely. This variation reflects individual country preferences, the Fund staff's requests that respond to particular surveillance or program needs, and other factors. A survey of a set of core statistical indicators reported to all area department country desks was undertaken as of end-February 2000 to assess the timeliness and frequency of the economic and financial data available to the Fund. This section reviews changes in the availability of statistical data and compares the results of this survey with those of the 1997 survey.10 Changes in data availability since the last survey have taken place against the background of the implementation of the Fund's standards initiatives and the increasing emphasis placed on the need for adequate and timely data to assess countries' vulnerability to financial crises. 69. In general, improvements in the provision of core indicators data since 1997 have been modest. With respect to the high-frequency indicators, Chart 1 suggests that the provision of this information to the Fund by most countries is relatively timely and there was little change since the last survey was conducted. Nevertheless, there were small increases in the number of countries reporting with lags of less than one month for exchange rates, broad money, and interest rates. At the same time, there was a pronounced rise in the number of countries that made their central bank balance sheets available with a lag of less than one month. For international reserves, there was no change in the number of countries making these data available with a lag of less than one month. This reflects the fact that the majority of countries provide reserves data to the Fund either daily, weekly, or monthly.
70. Concerning the low-frequency indicators, Chart 2 shows that the results were mixed in terms of timeliness. While the number of countries reporting data with lags of three months or less increased for all indicators, the overall government balance excepted, there was also a deterioration in timeliness for countries that were already late in reporting.11 Underlying these trends was a significant improvement in the timeliness of the availability of data for external debt and GDP/GNP, which is reflected in the sharp increase in the number of countries that now make these data available with lags of three months or less. It would appear that subscription to the SDDS acted as a spur to statistical development in some countries. A number of countries began compiling quarterly GDP series to meet the requirements of the SDDS.12 This improvement in periodicity was also reflected in the frequency of data provision to the Fund. 71. The periodicity of the data provided to the Fund is also important for monitoring developing economic and financial trends.13 For the high-frequency indicators, the February 2000 survey shows that data on exchange rates are compiled on a daily basis for roughly 80 percent of countries, whereas 21 percent of all countries compile data on international reserves on a daily basis (Table 1). This outcome compares favorably with the results from the 1997 survey, which indicate that 74 percent and 17 percent, respectively, of all countries compiled data on exchange rates and international reserves on a daily basis. 72. When analyzed across country groupings, the survey conducted in February 2000 shows that program countries, with few exceptions, disseminate the full range of high frequency indicators with the greatest periodicity. The exceptions concern advanced countries where the periodicity of some indicators is even more frequent. All of these countries make exchange rate data available on a daily basis and 89 percent of them disseminate daily interest rate data. The high periodicity of these indicators for program countries and advanced countries is consistent with the results of the survey conducted in June 1997.
73. In the case of low frequency indicators, differences in the periodicity of the data were more pronounced across country groups, a profile that has not changed significantly in recent years. For the overall government balance, nearly half of the countries in the advanced group produced these data only annually, whereas a much larger proportion of countries in other groups produced these data with higher frequencies (quarterly and monthly). Monthly periodicities for this indicator were the standard for developing, transition, and program countries. These results are comparable to those obtained from the June 1997 survey. 74. The periodicity of the GDP/GNP indicator, however, is less frequent for developing and program countries than for advanced and transition countries. For the former two groups of countries, these data are typically compiled on an annual basis which is in contrast to the quarterly periodicity for advanced and transition countries. As mentioned earlier, a number of SDDS subscribers and near-subscribers began to compile quarterly as well as annual GDP series to meet the requirements of the standard. 75. Concerning external debt data, the results of the February 2000 survey show noticeable cross country differences in the periodicity of these data as well as significant improvements since the previous survey. The majority of advanced and developing countries compile these data annually. In contrast, over 50 percent of the transition and program countries compiled these data with a periodicity equivalent to or in excess of a quarterly basis. The periodicity of these data has increased since 1997, a change that is largely attributable to improvements in transition and program countries. 76. To provide background for the recommendations in the present review regarding data on international reserves and foreign exchange liquidity, in March 2000 the staff conducted a survey of the content, timeliness, and periodicity of information that desk economists receive from country authorities. The survey was intended to provide a snapshot of current practices regarding the provision of the broad categories of data included in the reserves template for 42 countries and territories, principally emerging market countries.14 The results are summarized in Table 2. 77. The survey shows that a large majority of these countries already provide at least partial information to the Fund on the key components of the reserves template. The survey also shows that information on predetermined and contingent drains, in general, is provided at a lower frequency than reserve assets. In addition, concerns often exist about the accuracy and coverage of data on drains. This finding confirms statistical compilers' experience that the main difficulty of compiling information for the template in a timely manner relates to the availability of data on central government debt obligations (both predetermined and contingent)—obtaining information on reserve assets is generally less problematic—and to technical difficulties in correctly classifying data on drains.
78. The staff has reviewed the treatment of data provision to the Fund in summings up and staff reports for all Article IV consultations concluded in the period March 1999-February 2000. This review thus covers 131 countries and regions—22 industrial countries, 23 emerging-market economies, 16 transition economies, and 70 developing countries.15 79. This review confirms that considerable attention was devoted by both the Board and the staff to data provision to the Fund and related issues, such as data quality, in Article IV consultations. The treatment of data issues in Article IV consultation documents generally included a description of the availability and quality of macroeconomic statistics (national accounts, monetary, fiscal, and balance of payments data), information on progress made in addressing past statistical weaknesses, and recommendations to address remaining shortcomings. A detailed presentation of these elements can be found in the statistical appendices of Article IV consultation reports for most emerging-market countries, transition economies, developing countries, and for some industrial countries as summarized in Table 4. The main sections of these Article IV staff reports generally contained a summary of the most salient issues treated in statistical appendices. In a majority of these cases, key data weaknesses and proposed remedial actions were also referenced in summings-up. Treatment of data issues in consultation reports for some other industrial countries was more succinct, as those countries were perceived not to suffer from significant data limitations. 80. Four-fifths of Article IV staff reports and two-fifths of all summings up covered in the sample contained an explicit overall assessment of the adequacy for purposes of surveillance of data provided to the Fund (Table 5). These explicit assessments were more frequent in stand-alone Article IV consultation documents, where they were present in nearly all reports and a majority of summings up, than in documents for Article IV consultations combined with requests for, or reviews of, Fund arrangements. This difference may be explained, at least in part, by ongoing treatment of data weaknesses and corresponding remedial measures in the latter cases. 81. Approximately two-fifths of staff reports' assessments of data provided to the Fund reported that data deficiencies hampered surveillance, hindered surveillance, or limited economic analysis and policy formulation. There were no cases where the staff of the Executive Board expressed explicit concern that a member was not cooperating in providing data.
82. While data issues received substantial coverage in staff reports, their treatment was rarely integrated in the discussion of macroeconomic developments and policy issues. In particular, reports seldom attempted to draw attention to uncertainties regarding past economic developments, projections, or ensuing policy conclusions stemming from noted data deficiencies. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
