Ninth Review of the International Monetary Fund’s Data Standards Initiatives

The International Monetary Fund's Executive Board regularly reviews progress and developments under the Data Standards Initiatives.


The International Monetary Fund's Executive Board regularly reviews progress and developments under the Data Standards Initiatives.

The Data Standards Initiatives in Retropsect

1. The Data Standards Initiatives were launched after the financial crisis of 1994/95 on realization that data deficiencies and lack of transparency can contribute to market turmoil. Over time, the initiatives have proved valuable to the international community and country authorities, as demonstrated by near universal acceptance by IMF members; the willingness of many to commit—voluntarily—to observe high standards of data dissemination; and the recent establishment of a third and highest tier, the Special Data Dissemination Standard (SDDS) Plus in 2012. By end-March 2015, there are 113 participants in the General Data Dissemination System (GDDS), 63 subscribers of the SDDS, and eight adherents to the SDDS Plus (Figure 1 and Table 1); with only eight IMF members remaining outside of the Initiatives.1

Figure 1.
Figure 1.

Data Standards Initiatives: International Monetary Fund Membership Participation

(as of end-March 2015)

Citation: Policy Papers 2015, 042; 10.5089/9781498344777.007.A001

Source: Dissemination Standards Bulletin Board.
Table 1.

Data Standards Initiatives: Chronology1

article image
article image
Source: Dissemination Standards Bulletin Board.

Countries are listed according to the date of participation within a given year. Countries in bolded text indicate graduation from the GDDS.

2. The near universal acceptance of the Data Standards Initiatives reflects a number of factors. These include: (i) the “public good” aspect of data dissemination, given its criticality for the conduct of surveillance by the IMF as well as monitoring and analyses by market participants, the media, and the public; (ii) the pressure exerted by policy makers and analysts to address data gaps, which is essential for the formulation and implementation of economic policies aimed at promoting growth and macroeconomic stability; and (iii) strong support from the IMF’s Executive Board, including to ensure the initiatives’ ongoing relevance and effectiveness.

3. Each of the initiatives evolved differently:

  • The SDDS has played a dominant role in promoting the disciplined dissemination of timely and standardized data across a broad range of economic and financial statistics. When subscribing to the SDDS, a country commits to disseminate the data categories with prescribed coverage, periodicity, and timeliness on a National Summary Data Page (NSDP), and according to an Advance Release Calendar (ARC), while agreeing to non-observance procedures and sanctions that serve to enforce compliance. Policy makers, financial market participants, and other data users continue to point to the benefits of the SDDS in introducing the discipline of observing an advance release calendar, which eliminated uncertainty in data dissemination. Since 1998, subscribers have been required and encouraged to disseminate additional data categories in line with emerging needs. These data categories include the reserves data template, international investment position (IIP), and external debt; and the encouraged categories on financial soundness indicators, sectoral balance sheets, and government gross debt. In addition, more demanding requirements were introduced on periodicity and timeliness (for external debt and quarterly IIP data) and the provision of hyperlinks to times series data (Table 2).

    Table 2.

    Data Standards Initiatives: Main Modifications, 1998–2014

    article image
    Source: Reviews and Discussions of the IMF’s Data Standards Initiatives at

  • The new SDDS Plus builds on the progress achieved under the SDDS by introducing more rigorous requirements for disseminating a broader set of data, with no flexibility options for the nine additional datasets.2 Recognizing the challenges in compiling and disseminating the new data categories, however, the SDDS Plus allows for a transition period for implementation lasting through end-2019. Until that time, countries may adhere to the SDDS Plus by observing the requirements for at least five of the nine new data categories.3 Following the launch in November 2014, eight countries have adhered to the SDDS Plus—France, Germany, Italy, the Netherlands, Portugal, Spain, Sweden, and the United States. Their metadata and links to their NSDPs are posted on the IMF’s Dissemination Standards Bulletin Board (DSBB).

  • Notwithstanding some modifications to the framework introduced in 2008 at the time of the Seventh Review, the GDDS has remained largely unchanged (Table 2). At its core, the GDDS was designed as a developmental framework to assist countries with weak statistical capacity. It requires preparation of metadata and action plans to improve the compilation, coverage, periodicity, and timeliness of macroeconomic, financial, and selected socio-demographic data.

4. The evolution of the Data Standards Initiatives provides insights and lessons that could benefit future reforms.

  • Resources—particularly technical staff and IT—must be adequate: Progress under the Data Standards Initiatives to a large extent has been heavily influenced by the resources—human, financial, and technological—available to support statistical development, with advanced and emerging economies in a better position to undertake the investments needed to observe rigorous data standards. The slow rate of graduation from the GDDS to the SDDS—only 12 countries since the standards were established in the mid-1990s—reflect, among other things, the insufficiency of resources allocated in most low-income countries to the statistical function. The statistical development gap between the group of advanced and emerging countries and the group of low-income countries has widened with the establishment of the SDDS Plus.

  • Political will and commitment—as reflected by statistical legislation and effective inter-agency coordination—are essential for data transparency: Absent strong political support and a robust legal and institutional framework, there is little domestic incentive to prioritize data transparency. In such circumstances, participation in the GDDS has been characterized by a strong initial push involving officials of various agencies preparing the metadata and plans for improvement with facilitation by IMF missions and funding by donors. Once these initial tasks are accomplished, however, a slackening of efforts has been evidenced by stagnation in the metadata and plans for improvement (Table 3). Experience under the GDDS shows that the lack of incentives for implementing the plans for improvement—including to start regular data dissemination—and the absence of monitoring of progress, have often led to the loss of the initial impetus for statistical development.

    Table 3.

    GDDS Participants: Status of Metadata/Plans for Improvement Updates

    (number of participants, as of March 2015)

    article image
    Source: Data Standards Bulletin Board.

    Staff monitors the status of each participant’s metadata and plans for improvement and can track when this information is updated on the Data Standards Bulletin Board.

    Six economies that began participation in 2013-14 are excluded from this category (Bosnia and Herzegovina, Cook Islands, Marshall Islands, Myanmar, Palau, and Tuvalu).

  • The IMF’s policy dialogue with country authorities can be used to elevate the priority of quality data and disseminationin particular, in countries with weak capacity. This is in line with the 2014 Triennial Surveillance Review and the focus on closing data gaps. Success under the SDDS and SDDS Plus has demonstrated the critical role the IMF can play in leveraging its instruments—data standards initiatives, capacity building activities, and surveillance—to promote statistical development. As appropriate, Article IV consultation missions could include broad issues related with the Data Standards Initiatives—data adequacy, dissemination, and quality—in their discussions with the authorities, raising the profile of statistics in the policy dialogue.4

Recent Developments and Progress

A. Special Data Dissemination Standard

5. Two GDDS participants have graduated to the SDDS since 2012 (Mauritius and West Bank and Gaza) and five over the past five years. These low numbers are indicative of the challenges countries face for graduating from the GDDS to the SDDS, and the statistical-development gap between GDDS participants and those in the higher standards.

6. During the Eighth Review, the IMF Executive Board endorsed a few modifications to the SDDS, with the main change involving strengthened non-observance procedures. The timeline for actions by the staff and management, and the authorities in the event of non-observance were clearly spelled out under the new procedures.5 Close collaboration and consultation between staff and subscribers (through the SDDS Coordinator6) on emerging issues have so far, effectively pre-empted the need to invoke these procedures. Subscribers generally have strong capacity and observe the requirements diligently. As a consequence, at this time the standard is largely “self-policing,” requiring less staff resources for effective oversight. The commitment to the SDDS also is illustrated by the timely observance of the dissemination requirements for quarterly IIP data (effective September 2014, as prescribed by the Executive Board in 2010) by all but one subscriber.7

7. Subscribers are addressing—in collaboration with staff—the other modifications made to the standard earlier, but there is still work to do:

  • NSDPs should contain hyperlinks to the full set of prescribed data to provide users with quick access to more detailed and longer time series. As of end-March 2015, 43 of the 63 subscribers (and all eight SDDS Plus adherents) provided hyperlinks to time series for all data categories; the remaining 20 subscribers have one or more data categories that do not have such links. While there is no timeframe, staff is working with the authorities to ensure the NSDPs have the required hyperlinks as soon as feasible.

  • Clarity about the methodological foundations as well as any deviations from internationally accepted statistical methodologies is important for transparency and for facilitating appropriate cross-country comparisons. Subscribers were asked for the metadata to reflect full methodological disclosure. Through March 2015, 53 of the 63 have complied, with all eight of the SDDS Plus adherents meeting this requirement.8

  • Modifications also encouraged the dissemination of sectoral balance sheets and general government total gross debt in nominal values, as well as seven financial soundness indicators (FSIs) in 2010.9 However, no SDDS subscriber disseminates sectoral balance sheet data; only one disseminates general government total gross debt in nominal values; and 56 of the 63 provide the seven FSIs. In comparison, all eight SDDS Plus adherents disseminate sectoral balance sheets data; six disseminate data on general government total gross debt in nominal values; and three disseminate the FSIs (including residential real estate prices, which are not included under the SDDS).

8. Given the generally good track record of subscribers in the observance of the SDDS requirements, lighter monitoring is now warranted. In this regard, staff plans to take a more risk-based approach to the monitoring of the SDDS. The priority for this standard is moving those subscribers with systemically important financial systems to the SDDS Plus, assisting GDDS participants to graduate to the SDDS, and consolidating the modifications introduced in previous reviews.

9. Staff continues to promote graduation to the SDDS, especially among members that demonstrate strong commitment to high data standards and those with large economies. The former group includes countries that are working closely with staff to implement plans to meet the SDDS requirements in the near term (for example, Albania, Bolivia, Guatemala, Montenegro, Panama, Serbia, Sri Lanka, and Seychelles) and those that are building good statistical capacity but might only meet the requirements over the medium term (for example, Botswana, Bangladesh, Namibia, Mongolia, Pakistan, and Uganda).10 Staff is also working closely with China and Saudi Arabia, the only two G20 economies that are not subscribers to the SDDS (or adherents to the SDDS Plus). Following the November 2014 announcement by President Xi Jinping of his government’s intention to subscribe to the SDDS, a mission assessed China’s data compilation and dissemination practices against the standard’s requirements. Staff understands that China could subscribe by end-2015. At the request of the Saudi Arabian authorities a mission visited Riyadh in March 2013 to assess the country’s practices, and the authorities have been working to meet the SDDS requirements, although the timeframe for subscription has not yet been conveyed to staff.

B. Special Data Dissemination Standard Plus

10. The SDDS Plus was established in 2012 in the context of the Eighth Review, as its third and highest tier. It features rigorous standards for nine additional data categories beyond those in the SDDS,11 and complements the G-20 Data Gaps Initiative (DGI), which the IMF is actively promoting.12 Given the consistency between the information needs of the DGI and the SDDS Plus, as countries make progress under the DGI, they also advance towards adherence to the SDDS Plus.

11. The launching of the SDDS Plus took longer than initially envisaged owing to a number of factors. While some of the challenges for adherence are country-specific, many are common to all (prospective) adherents including: (i) tight budget constraints; (ii) trade-offs in allocating scarce resources either to meet the SDDS Plus’ demanding standards or competing statistical priorities;13 (iii) uncertainty about the capacity to meet the requirements by the end of the transition period, in particular, for data categories that are only partially or not yet compiled (e.g., debt securities); and (iv) technological challenges relating to the use of the Statistical Data and Metadata Exchange (SDMX)—the platform supporting the SDDS Plus.14

12. Considerable efforts were required to facilitate adherence, including bilateral consultations, preparation of Guides, and tweaks to the framework (Box 1). The tweaks to the framework made in March 201415 involved lengthening the timeliness of three data categories (sectoral balance sheets, other financial corporations survey, and debt securities) from three to four-months after the reference period.16 A change also was made to one of the FSIs (ratio of liquid assets to short-term liabilities) to ensure consistency with updated methodologies under the Basel Accords. In addition, staff worked closely with several authorities to revamp the NSDP and incorporate an SDMX platform. This revamped NSDP is user friendly, machine readable, and removes the need for manual updating; however, it requires an upfront investment that will yield cost-saving.

SDDS Plus: Outreach and Consultation

Since the IMF Executive Board approved the SDDS Plus staff has undertaken a number of outreach events and worked closely with authorities to facilitate adherence. Activities include:

  • Workshops: Two SDDS Plus workshops were held during: (i) September 26–28, 2012 (Washington, D.C.) for delegations from 23 countries; and (ii) November 12–13, 2014 (Frankfurt) for 47 delegates from 29 countries. During the Washington workshop staff engaged delegations on the key elements of SDDS Plus, including presentations on the nine new data categories and exchanged views on the information-technology platform to support the initiative. During the Frankfurt event staff elaborated on the technical requirements for using SDMX, and delegations engaged in peer learning, with German, Spanish, and the United States officials sharing their hands-on experience with implementation and country-specific approaches.

  • Correspondence: Staff invited expressions of interest in the SDDS Plus and offered technical support to all 71 SDDS subscribers in 2012. To date, 23 subscribers have expressed interest in transitioning to the SDDS Plus.

  • Meetings: Staff made presentations on the SDDS Plus to the European Commission’s Committee on Monetary, Financial, and Balance of Payments (July 2013, January 2014, July 2014), Bank of France (June 2013), and a Regional Meeting on the SDDS Plus in Frankfurt (May 2014). During these meetings, staff provided clarification on the new data categories as well as on steps for adherence.

  • Bilateral consultations: SDDS Plus consultation missions were undertaken to the eight countries in the first cluster of adherents and to Austria, Chile, Czech Republic, and Japan.

  • Publication of guidance notes: These included the (i) Special Data Dissemination Standard Plus: Guide for Adherents and Users (Users Guide); (ii) National Summary Data Page: Technical Implementation Guide (Technical Guide); and (iii) Clarification Notes on the treatment of Nonautonomous Unfunded Government Employee Pension Schemes in the General Government and Sectoral Balance Sheets Data Categories and Other Accounts Payable in General Government Total Gross Debt (in nominal values) and Sectoral Balance Sheets Data (Clarification Notes).

13. The launch of the SDDS Plus took place in November 2014 and the first cluster of countries met the requirements in February 2015. This cluster includes France, Germany, Italy, the Netherlands, Portugal, Spain, Sweden, and the United States and their metadata as well as links to their respective NSDPs are posted on the DSBB.

14. Staff does not envisage further modifications to the SDDS Plus considering that implementation is in its early stages and more time will be needed to assess developments. Nevertheless, staff will remain open to feedback from current and prospective adherents on potential problem areas and suggestions on how implementation might be accelerated.

15. In line with one of the impetuses behind the SDDS Plus, the priority for the period ahead will be to promote adherence by economies with systemically important financial systems that play a leading role in international capital markets. These economies are those for which the IMF Executive Board has mandated financial stability assessments under the Financial Sector Assessment Program (FSAP) every five years.17 Under the current resource envelope, STA has the capacity to assist the adherence of about 5–6 countries per year. Experience with the first cluster of countries indicates that—once the country is producing the required data—adherence would take about 3–4 months of close collaboration between staff and the authorities. This will involve primarily transitioning all data categories—including those under the SDDS—to the revamped NSDP using SDMX.

C. General Data Dissemination System

16. Since the Eighth Review in 2012, 13 countries became participants in the GDDS (Table 1). These include nine from Asia and the Pacific, one each from Europe and the Middle East and Central Asia, and two from sub-Saharan Africa. As of end-March 2015, there are 113 GDDS participants.18

Modifications during the Seventh Review

17. At the time of the Seventh Review in 2008, the IMF Executive Board endorsed the only substantive modifications to the GDDS since its establishment (Table 2). A primary motivation for the changes was to place stronger emphasis on data dissemination, with the aim of facilitating graduation to the SDDS. To this end, data categories were aligned with those of the SDDS; participants were encouraged to develop NSDPs and Advance Release Calendars (ARCs) to publish data in an internationally recognized and standardized format,19 and participants also were expected to address data quality issues (e.g., compilation and coverage) through regular updates of plans for statistical improvement.

18. The expectation was that the 2008 modifications would re-invigorate the GDDS, with participants intensifying efforts to develop their statistical systems. Indeed, staff thought that the changes would: (i) catalyze greater resources to expand the scope and coverage of macroeconomic and financial statistics; (ii) encourage countries to adopt standardized dissemination practices, particularly NSDPs and ARCs; (iii) promote adoption of modern compilation techniques and new source data (e.g., surveys to compile more detailed external sector data); and (iv) raise the importance of data dissemination for transparency, sensitizing the authorities about the need for publication of the full range of statistics, including international reserves.

19. Insufficient progress, however, has been made in implementing the changes called for under the Seventh Review. While data categories were aligned to the SDDS and a few participants developed both NSDPs and ARCs (14 participants in sub-Saharan Africa) (Box 2 and Appendix I), for the majority of participants metadata are rarely updated and data dissemination and plans for improvement have largely not materialized. During the past three years, only 42 of the 106 GDDS participants (excluding those that joined only in the last two years) have modified this information, while 50 have not done so in over seven years, including 29 that have never revised this information (Table 3 and Appendix I). Only 12 GDDS participants have graduated to the SDDS after nearly two decades of the SDDS’ existence (in chronological order: Kazakhstan, Armenia, and Bulgaria (2003); Kyrgyz Republic (2004); Romania (2005); Moldova (2006); Malta (2009); Jordan and Georgia (2010); Macedonia, FYR (2011); Mauritius and West Bank and Gaza (2012)).

Donor Support for the GDDS

The promotion of data dissemination under the GDDS has been strongly supported by the United Kingdom’s Department for International Development (DFID) and Japan, under the Japan Administered Account for Selected IMF Activities (JSA).

A DFID project (2010–15) has provided financing of US$7.5 million for 10 modules, including one on the Enhanced Data Dissemination Initiative (EDDI) designed to assist 25 Anglophone countries in sub-Saharan Africa to build dissemination capacity. Technical assistance under the project has facilitated preparation of NSDPs and ARCs, especially in countries seeking graduation to the SDDS. Under the project, Burundi has become a GDDS participant, 11 of the countries covered by the project created NSDPs and 13 produced ARCs, and Mauritius was the first of the countries covered by the project to graduate to the SDDS (February 2012). In February 2015, DFID approved the second phase of the EDDI for a five-year period beginning in April 2015 and amounting to about US$9.5 million.

Japan has financed two projects. The first, initiated in 2011 provides US$1.3 million to assist 13 economies (the Cook Islands, Iran, Lao PDR, Marshall Islands, Micronesia, Myanmar, Palau, Papua New Guinea, Samoa, Timor-Leste, Turkmenistan, Tuvalu, and Uzbekistan) to participate in the GDDS, mainly through technical assistance and regional workshops. To date, the project has led to 11 new GDDS participants. The second project, initiated in May 2014 provides US$2.1 million to assist 12 countries in Southeast Asia and the Pacific (Bangladesh, Bhutan, Brunei Darussalam, Cambodia, Mongolia, Myanmar, Nepal, Papua New Guinea, Samoa, Sri Lanka, Timor-Leste, and Vietnam) to disseminate data and graduate to the SDDS.

20. Consultation and outreach with GDDS participants and other stakeholders suggested a number of factors behind insufficient progress in data dissemination. A main factor cited in the feedback is the stagnation of the current system, which embraces the notion that development of metadata and plans for statistical improvement should provide sufficient impetus to disseminate data and set countries on a path to higher standards. Other factors include inadequate financial and human resources and technical capacity, especially in sub-Saharan Africa and fragile states.

21. GDDS stakeholders also stressed the link between incentives and political support for statistics, urging a stronger statistical dimension in the IMF’s surveillance dialogue (Box 3). They were of the view that a high-level commitment to disciplined data dissemination is a critical factor to motivate improvement in data quality. In particular, senior statistical officials—from central banks, ministries of finance, and national statistical offices—argued that sharpening the focus of the GDDS on disciplined data dissemination will be essential to mobilize the resources and political support needed to set countries on the path towards data quality improvement and SDDS subscription. In this connection, raising the profile of data quality and dissemination weaknesses in Article IV consultations was seen as crucial to: (i) elevate data issues in internal/domestic policy discussions; (ii) highlight resource and capacity constraints; and (iii) strengthen coordination and collaboration among the members of the national statistical system.

Consultations with GDDS Stakeholders

For assessing progress and seeking views on options for enhancing the GDDS, staff consulted with GDDS participants, the private sector (selected investment banks and three rating agencies), development partners (World Bank, African Development Bank, Inter-American Development Bank), the United Nations, and the IMF Executive Board during an informal session in February 2014.

Consultations with GDDS participants included regional workshops in sub-Saharan Africa in July 2014 (Uganda, 16 countries), the Middle East in June 2014 (Oman, 6 countries), Southeast Asia in April 2014 (Fiji, 8 countries) and August 2014 (Vietnam, 8 countries); and visits to Senegal, Cameroon, and St. Kitts and Nevis for discussions with the authorities and regional institutions (BCEAO, BEAC, and the Eastern Caribbean Central Bank).

Stakeholders identified weaknesses in the GDDS framework and made recommendations that staff has taken onboard in proposing reforms (Section 3 in the main text). Prominent among the weaknesses identified were: (i) lack of incentives for policy makers to allocate sufficient resources for data dissemination under the GDDS; (ii) absence of benchmarks, external monitoring, and peer review; and (iii) inadequate coordination of statistical capacity development efforts at both the national and international levels.

Stakeholders generally agreed that the GDDS framework should be enhanced along three dimensions:

  • Stronger coordination and collaboration within countries and among technical assistance/training providers. Insufficient inter-agency coordination and collaboration within the national statistical establishment was cited as a significant barrier to progress by a majority of stakeholders;

  • More effective leveraging of the IMF’s surveillance activities to elevate statistical issues to the level of policy makers. Stakeholders advised staff to closely link the Data Standards Initiatives with surveillance activities, which could decisively strengthen the constituency for better data and more resources, thus creating a virtuous circle for reform; and

  • Development of benchmarks to monitor and assess progress, and facilitate peer review, which likely will promote domestic and international interest in data dissemination. In this regard, the majority of stakeholders viewed data dissemination as a critical requirement for improving data quality.

Why has graduation to the SDDS been so slow?

22. The SDDS was established under the premise that it could help promote transparency for countries aspiring to access international capital markets. In the event, a study conducted on the basis of about a decade of experience, found a potential benefit from SDDS subscription through lower borrowing costs in international capital markets, estimated at about 20–50 basis points.20

23. Since the global financial crisis, however, accommodative monetary conditions and the “search for yield” appear to have softened investors’ demands as regards data transparency. Many GDDS participants have accessed international capital markets or received sovereign credit ratings (17 GDDS participants during 2009–14, as indicated in Table 4 and Appendix 1) without subscribing to the SDDS, which suggests that the potential “fiscal benefit” has diminished since 2008 (Box 4).

Table 4.

GDDS Participants with Sovereign Credit Ratings and Access to Capital Markets, by Region

(as of December 20141)

article image
Sources: IMF, Monetary and Capital Markets and Statistics Departments.

First access after 2004; proceeds cover government financing needs (excluding debt restructuring) and the minimum issue amount is US$200 million.

First access after 2007-08 (see Appendix 1).

Data Dissemination—A Public Good

Data dissemination brings a broad range of benefits for the country and the global community. An earlier review of the IMF’s Data Standards Initiatives identified the availability of timely economic and financial statistics as a global public good, critical for the proper functioning of financial markets and global financial stability.21 This public good is essential for providing information to markets, but even more because it puts a constraint on what policy makers can do. Fischer (2002) noted that transparency promotes interactions with the outside world, as information is presented, providers must interact and listen to what outsiders are saying, and take information in; it also can strengthen the effectiveness of IMF surveillance.22

Data dissemination can help countries meet diverse needs. Public access to data can promote data quality because only by making data available, it is possible to assess and improve information, the soundness of methodologies, and accuracy and reliability; all of which support policy analysis and decisions. Benefits also accrue from international comparisons and peer review of data.

24. A number of GDDS participants are strong candidates for SDDS subscription. Alexander and others (2008) using 2005 data observed that SDDS subscribers tended to have relatively high per capita GDP, with nearly all subscribers above US$2,000 (2005) reflecting the likelihood that statistical capacity building requires significant resources.23 Using this indicative income threshold at 2013 prices (approximately US$2,600, based on a 5 percent annual inflation rate), about half of the current GDDS participants might have been expected to move to the SDDS.

25. Broadening this analysis to include other dimensions that ought to be important for building statistical capacity—in particular, the strength of institutions and human development—leads to a similar magnitude of potential SDDS subscribers. Real per capita GDP, strength of institutions measured by the Polity2 score, and human development captured by the UN’s Human Development Index can be combined to estimate the distance between the average GDDS participant and the average SDDS subscriber. Results suggest that about 50 GDDS participants share similar income and institutional and human development characteristics as the average SDDS subscriber (the detailed analysis for the distance index is provided in Appendix II).

26. Identifying possible GDDS candidates for graduation offers useful insights, although is only indicative. The long experience under the Data Standards Initiatives suggests that many factors play a role in statistical development, including per capita income (as a proxy for financial constraints), strength of institutions, and political commitment. Indeed, without strong political commitment the other factors may lose importance. Consequently, strengthening the policy dialogue and incentives for data dissemination under the GDDS is critical for progress.

A Proposal to Enhance the GDDS

27. Staff proposes refocusing the GDDS on the publication of data essential for surveillance by the IMF and markets, while leveraging the Article IV consultation policy dialogue to direct the authorities’ attention to progress made along a three-stage path towards SDDS subscription. In contrast to the present framework under which GDDS countries are encouraged to disseminate certain data with no monitoring, under the proposed enhancement staff (STA) would prepare progress reports on data dissemination for discussion with the authorities by Article IV consultation missions (see also paragraph 42). Under the proposal, the GDDS would be “enhanced” but not replaced by a new standard, in line with the feedback from the outreach consultations.

28. There are four elements to the proposed enhanced—GDDS (“e-GDDS”):

  • a revision to the encouraged data categories;

  • a renewed focus on disseminating data in a standardized format;

  • annual monitoring of progress and developments; and

  • leveraging surveillance activities to support statistical improvement.

29. The e-GDDS would apply to all current and future GDDS participants.24 Including all GDDS countries in the e-GDDS would serve to encourage all participants to implement the elements of the new framework. Indeed, staff would be able to monitor progress in statistical development (Section C) only when countries disseminate data through an NSDP (Section B).

A. Revised Data Categories

30. Staff proposes to align the GDDS data categories with those required under Article VIII, Section 5, as listed in the Table of Common Indicators Required for Surveillance (TCIRS). This alignment would result in 15 encouraged data categories, which provide the basis for the discussions routinely held by country teams with the authorities, creating synergies with the surveillance activities. While the authorities are required to provide to the IMF the TCIRS data, under the e-GDDS the authorities also would voluntarily commit to publish the same data in a disciplined manner—in accordance with an advance release calendar.25 This alignment would have several advantages:

  • The TCIRS is well known to IMF country teams and the authorities, and the data categories have been endorsed by the IMF Executive Board as the core indicators needed to undertake surveillance pursuant to Article VIII, Section 5 of the Articles of Agreement (Box 5). Staff is required to report on provision and overall adequacy of the TCIRS data in Article IV consultation staff reports (cf. the Statistical Issues Appendix) but the authorities are not obligated in any way to disseminate this information, other than to the IMF. Closely aligning the e-GDDS to the TCIRS would help to improve ownership of these data, raise their profile, and lead to support for improvement.

    Table of Common Indicators Required for Surveillance: Background

    In 1995 the IMF Executive Board endorsed a “basic set of data common to almost all countries that reflect the core areas of macroeconomic statistics,” to support the IMF’s surveillance mandate (SUR/95/180; SM/95/180, Attachment II). The data included 10 macroeconomic and financial variables which were listed in the Core Statistical Indicators Table (exchange rates, international reserves, reserve or base money, broad money, interest rates, consumer price index, exports/imports, external current account balance, overall fiscal balance, and GDP/GNP).

    Over time, the set of data has been expanded (external debt was added in 2000 and IIP in 2010) and the Core Statistical Indicators Table was replaced by the Table of Common Indicators Required for Surveillance in 2004 (TCIRS; SM/04/56, Box 5 and Table 1). At that time new information was added covering data frequency and timeliness, and other aspects of data quality.

    Box 5 Table 1.

    Table of Common Indicators Required for Surveillance

    article image

    Any reserve assets that are pledged or otherwise encumbered should be specified separately. Also, data should comprise short-term liabilities linked to a foreign currency but settled by other means as well as the notional values of financial derivatives to pay and to receive foreign currency, including those linked to a foreign currency but settled by other means.

    Both market-based and officially-determined, including discount rates, money market rates, rates on treasury bills, notes and bonds.

    Foreign, domestic bank, and domestic nonbank financing.

    The general government consists of the central government (budgetary funds, extra budgetary funds, and social security funds) and state and local governments.

    Including currency and maturity composition.

    Includes external gross financial assets and liabilitiy positions vis-à-vis nonresidents.

    Daily (D); weekly (W); monthly (M); quarterly (Q); annually (A); irregular (I); and not available (NA).

    Reflects the assessment provided in the data ROSC or the Substantive Update (published in July 2010, and based on the findings of the mission that took place during December 09-22, 2009) for the dataset corresponding to the variable in each row. The assessment indicates whether international standards concerning concepts and definitions, scope, classification/sectorization, and basis for recording are fully observed (O); largely observed (LO); largely not observed (LNO); not observed (NO); and not available (NA).

    Same as footnote 7, except referring to international standards concerning source data, statistical techniques, assessment and validation of source data, and revision studies.

  • Moreover, as the TCIRS plays a central role in the surveillance activities, this alignment would integrate and leverage both instruments and create synergies. This also would help (i) staff’s assessment of data shortcomings; (ii) identification of remedial measures and capacity development needs; and (iii) clarification of the areas of responsibility for data issues among area departments, other functional departments, and STA in support of dissemination (cf. Section II.C. in SM/13/155).

  • Closely aligning the e-GDDS data categories with those of the TCIRS also would help define thresholds for a core set of data to be disseminated (paragraph 34), in the same spirit as the IMF Executive Board has endorsed a core set of data needed to support surveillance. These thresholds would serve to monitor progress under the e-GDDS, addressing the concerns raised in the consultations with stakeholders on the lack of benchmarks to assess statistical development.

31. The proposed data categories and their coverage, periodicity, and timeliness are presented in Table 5 and detailed below:

Table 5.

e-GDDS: Proposed Data Coverage, Periodicity and Timeliness

Macroeconomic and Financial Sectors and Socio-demographic Data

article image
article image

Based on BPM6 categories; BPM5 basis data should be presented in equivalent detail.

  • The e-GDDS would cover 15 data categories in the real, fiscal, monetary and financial, and external sectors, mirroring the TCIRS (Box 5 and Table 5). Participants also would be encouraged to provide statistics for three supplementary data categories—production index, labor market indicators, and producer price index—that are not part of the TCIRS but are included in the SDDS.26 Participants would continue to be required to provide metadata, including for the supplementary data categories.

  • The periodicity and timeliness of the e-GDDS would be much less stringent than the SDDS (Table 7), except for national accounts and balance of payments, where they are proposed to be the same. The more stringent requirements for national accounts and balance of payments would reflect (i) the importance of these data for surveillance; and (ii) the fact that many GDDS participants (about one quarter) are already publishing these data as proposed (Table 8).

Table 6.

SDDS Subscribers: Use of Flexibility Options

(number of subscribers per data category)

article image
Source: Dissemination Standards Bulletin Board.

The SDDS allows subscribers to use flexibility options for up to two data categories. These flexibility options allow them to disseminate data with a periodicity or timeliness (or both) “less” than prescribed under the SDDS. Flexibility options are not permitted for official reserve assets, data template on international reserves and foreing currency liquidity) and external debt. Fifty-four SDDS subcribers take advantage of the flexibility option(s).

n.a. implies no flexibility option is available for periodicity and/or timeliness.

Table 7.

SDDS and e-GDDS: Comparison of Coverage, Periodicity, and Timeliness

article image

Dissemination of supplementary data categories is also encouraged, if available.

Participants in the e-GDDS would not be required to disseminate the Template on International Reserves.

Table 8.

Data Compilation of Proposed e-GDDS Data Categories—Current Practice

article image
Source: Dissemination Standards Bulletin Board.

32. In addition, the e-GDDS would encourage the dissemination of financial soundness indicators (FSIs), as in the SDDS. This set of indicators includes: regulatory Tier 1 capital to

risk-weighted assets; regulatory Tier 1 capital to assets; nonperforming loans net of provisions to capital; nonperforming loans to total gross loans; return on assets; liquid assets to short-term liabilities; and net open position in foreign exchange to capital. Dissemination of these data would be consistent with the IMF’s efforts to strengthen financial sector surveillance to better detect systemic risks and add a forward-looking element to the e-GDDS in anticipation of demands for more financial data in the period ahead. Notably, of the 100 countries submitting FSI data to the IMF on a regular basis, 34 are GDDS participants (Table 9).

Table 9.

GDDS Participants: Reporters of Financial Soundness Indicators

(by region)

article image
Source: Financial Soundness Indicators Database, IMF.

B. Data Dissemination

33. Under the e-GDDS, participants would be encouraged to disseminate the 15 data categories, with monitoring by staff. Publication in an easily accessible and internationally-comparable format is the missing element in the current framework. Staff monitoring will inform the international community on compliance with the ARC as well as on the progress in transiting to the SDDS through a three-stage path.

34. The proposal defines three thresholds in the path to readiness for subscription to the SDDS (Table 10). By introducing these thresholds, the proposal offers a vision of statistical development as a continuum, moving away from the view that advancement involves a “jump” to the SDDS. In the baseline (current system), participants disseminate metadata, plans for improvement, and none or some of the encouraged data categories. GDDS participants without an NSDP will be considered as in the baseline until they meet the requirements of the first threshold.

Table 10.
Table 10.

e-GDDS: Graphical Presentation of Monitoring Thresholds

Citation: Policy Papers 2015, 042; 10.5089/9781498344777.007.A001

Note: TCIRS refers to the Table of Common Indicators Required for Surveillance, NSDP to the National Summary Data Page, and ARC to the Advance Release Calendar.
  • Threshold one. Meeting the first threshold would involve disseminating—at least quarterly—through an NSDP the 15 data categories according to the coverage, periodicity, and timeliness stated in a country’s metadata, some of which may fall short of the e-GDDS framework.

  • Threshold two. Meeting the second threshold would involve disseminating the 15 data categories according to the coverage, periodicity, and timeliness recommended under the e-GDDS framework, with monthly updating of the NSDP.

  • Threshold three. Meeting the third threshold would involve a record of disseminating data according to, or better than the coverage, periodicity, and timeliness recommended under the e-GDDS framework and in line with an ARC, while maintaining an up-to-date NSDP. Participants meeting this threshold are nearing the requirements of the SDDS, although strong political commitment would be needed to reach this standard.

35. While monitored dissemination will not cover socio-demographic data, participants would be able to use the NSDP to publish such data. Currently the framework covers a few socio-demographic data categories (population, health, poverty, and education), but these indicators are generally not compiled by the authorities. Rather than re-dissemination of these data, which are available on the websites of the United Nations and World Bank, many stakeholders have suggested—including the staffs of the World Bank and the United Nations—for the e-GDDS platform to allow dissemination of socio-demographic indicators to be selected by the authorities. These could include those linked to the Sustainable Development Goals (SDGs). Such a selection could be made in consultation with specialized United Nations agencies and multilateral/regional development banks. The use of the NSDP for the dissemination of the SDGs has considerable promise for monitoring progress under the Financing for Development initiative.

36. The e-GDDS decision would be revised to reflect the changes to the framework. In contrast to the SDDS and SDDS Plus, it is not proposed that non-observance procedures be established for participants unable to disseminate the encouraged data. The enhancement aims at stronger incentives to disseminate data, rather than penalizing poor performance or weak capacity.27

37. Some countries will not be able to disseminate all of the e-GDDS data because of capacity and other constraints. Although dissemination is encouraged rather than required, the proposal is for staff to discuss the constraints to dissemination, in the context of the Article IV consultation, and to report findings, the authorities’ views, and remedial plans in the Article IV staff report (see Section C and Box 7 for further elaboration).28 Raising dissemination issues in this context is expected to rally support—internally and externally—for better data. As the e-GDDS data categories would be aligned with the data required for surveillance, enriching the consultation discussions in this manner should not increase the mission’s burden or require additional resources for country teams. While enhancing the policy dialogue on statistics should not be taxing for the authorities, redressing the barriers to statistical development—in particular for dissemination—will require adjusting budgetary priorities.

38. Under the e-GDDS, participants would disseminate data through an NSDP modeled on the information-technology platform recently established under the SDDS Plus. The central feature of this platform is the use of SDMX as the enabler. SDMX is a data exchange platform based on a global standard open format that offers many advantages to support the NSDP: it provides push/pull capabilities to facilitate internal and external data exchange and coordination; it is machine readable, which will help improve efficiency and reduce (human) errors; and it is easily accessible for users. The proposed NSDP would have a relatively simple web layout to reduce cost. Customized to the authorities’ preferences, the NSDP would feature columns indicating the GDDS data categories, hyperlinks to the authorities’ dissemination sites (for example, the websites of the national statistical office, central bank, or ministry of finance), and hyperlinks to the electronic SDMX data files.

39. Participants in the e-GDDS may also disseminate the encouraged data through an NSDP based on an open-data platform, or ODP. This electronic platform shares similar advantages with an SDMX-based NSDP (e.g., machine-readable, push/pull capabilities; data exchange; enhanced coordination) with the difference that there is an initial investment cost and an annual fee for service, which would be paid to the ODP provider (Box 6). Some of these costs could be offset by assistance from development banks and donors, several of whom have indicated strong support for the e-GDDS and the ODP initiative. Indeed this option is gaining traction worldwide, with support from the United Nations and the World Bank and regional development banks (African, Asian, Inter-American, and Islamic).

The Open Data Platform

The Open Data Platform (ODP) is a data dissemination platform provided by a (private) third party through a centrally-hosted portal (in the cloud, i.e., offsite through an externally-managed server). Using the ODP, an agency can own, create, and customize its own web-based NSDP.

Staff has been partnering with the African Development Bank (AfDB) to train officials in selected African countries in the use of ODPs for sharing data with the Fund, the AfDB, and other international organizations. Under the e-GDDS, the Fund will intensify its collaboration with the AfDB, including to use the ODPs already established in the countries in the region (with funding from the AfDB) to create the NSDPs.

Access to this service involves an initial one-time setup cost and annual fees (about US$10,000 and US$3,000 in the case of the AfDB-IMF project). The data producing agency loads data in the ODP using Excel-based files, and the data can be automatically converted into tables, charts, through a “dashboard” (i.e., internet browser), as well as in SDMX machine-readable formats, as under the SDDS Plus framework.

40. Progress under the e-GDDS would be assessed at the time of the next review of the Data Standards Initiatives. With the current resource envelope, STA would be able to assist 10–15 countries to implement the e-GDDS each year (Section E). IMF technical assistance to e-GDDS countries would involve re-calibration of priorities in consultation with country teams and donors. In the case of a large demand for assistance, STA staff would work with area departments to establish priorities as part of the consultative Resource Allocation Program that would aim to reflect countries’ demand (paragraph 46).

C. Monitoring and Reporting

41. Staff will prepare an annual report and a brief semiannual update. The main information source would be the monitoring of NSDPs. Once participants disseminate via the NSDP, STA staff would be able to assess progress against the core data categories, making reference to periodicity and timeliness. For each participant with an NSDP, STA staff also would make a judgment as to what threshold for dissemination of the data the participant meets. A table on the status (prepared by STA), for each country with an NSDP would be posted on the DSBB and a short summary on data dissemination issues also would be included in the Statistical Issues Appendix (SIA) accompanying each Article IV staff report (Box 7).29

Monitoring Under the e-GDDS

Every Article IV report has a Statistical Issues Appendix (SIA) that provides information necessary to form a view on the adequacy of data provision to the Fund for surveillance purposes (see SM/13/155, section C and Appendix II).

In addition to the TCIRS, the SIA includes the staff’s assessment of data shortcomings that have implications for surveillance. Also, it contains information on (i) a country’s data adequacy classification (i.e., adequate for surveillance, “A”; has some shortcomings but is broadly adequate for surveillance, “B”; and has serious shortcomings that significantly hamper surveillance, “C”); (ii) date of the latest data ROSC, if applicable; and (iii) adherence to the SDDS Plus, subscription to the SDDS, or participation in the GDDS.

Under the e-GDDS, country teams would elaborate further on the assessment of data adequacy informed by a discussion on data dissemination, with respect to the e-GDDS data categories. The summary text in the SIA would be augmented to reflect the e-GDDS participant’s dissemination practices (Box 7 Table 1 shaded area). These practices also would be discussed in STA staff’s assessment on Current Dissemination Practices Compared with the e-GDDS data categories (Box 7 Table 2). This latter assessment would be updated (semi-annually) by STA and posted on each e-GDDS’s participant’s country page on the DSBB.

Table 1.

[Country]—Statistical Issues Appendix

As of [Date]

article image
Table 2.

[Country]: Current Dissemination Practices Compared with the e-GDDS Framework

article image

D. International Support for the e-GDDS

42. Support by the international community will be important for the success of the proposed reforms. This support could include three dimensions:

  • First, aligning the e-GDDS data categories with those required for IMF surveillance likely will increase interest in country performance as regards data dissemination, including by rating agencies and other market participants. By integrating elements of Data Provision to the Fund for Surveillance Purposes and the Data Standards Initiatives, the profile of dissemination would be raised with policy makers.

  • Second, new ability to monitor progress in data dissemination based on transit from one threshold to another will afford a new way to assess capacity development needs, with the thresholds serving to guide decisions on technical assistance and donor support. With regard to technical assistance, this would involve setting priorities to improve the core data categories required for surveillance as well as dissemination.30 For relatively weak-capacity countries, this could imply assistance to strengthen plans for improvement; and for others, targeting assistance on specific data categories. With regard to training, this would involve taking advantage of regional synergies or economies of scale in providing courses to a range of countries facing similar challenges and issues.

  • Third, the necessary establishment/strengthening of the dissemination platform, in close collaboration with regional development banks will serve to facilitate publication of other data important for ascertaining social and demographic conditions, including those related with the SDGs.

43. Subject to resource availability, staff would stand ready to provide technical assistance in support of the e-GDDS, including as regards information technology. It is expected that many participants would not, however, require such assistance. In particular, those that are well advanced in data dissemination (e.g., 19 participants that have had SDDS assessments and are making progress toward that standard), have strong capacity, and are already developing ODPs. In close collaboration with the AfDB, staff has provided technical assistance to 20 countries in sub-Saharan Africa; and donor financing will allowed the Fund to continue this work in Africa, the Middle East and Central Asia, and in Asia and the Pacific (Box 2).

44. Some participants, however, will likely require substantial capacity development. On current resources, STA could support 10–15 participants each year to develop NSDPs. To optimize the use of scarce resources and meet demand, staff would give high priority to participants that: (i) demonstrate strong ownership of statistical development, as evidenced by regular and timely updates of their metadata and plans for improvement; (ii) have or desire access to international capital markets; and (iii) regularly submit data to the IMF’s International Financial Statistics (IFS), thus demonstrating a commitment to dissemination.31 As noted earlier, broad-based support for the e-GDDS from donors and regional development banks could also help alleviate these resource constraints.

E. Resource Implications

45. The e-GDDS would have resource implications for STA over the medium-term. STA would accommodate these resource requirements through reducing lower priority activities and redirecting capacity building to priority e-GDDS countries. It is not expected that the e-GDDS will add much additional burden on area department country teams. Under the current framework for surveillance, area department country teams are required to discuss data issues with the authorities. The e-GDDS proposal would only require area department country teams—with STA assistance—to discuss with the authorities dissemination issues on data already included under the current framework for surveillance. As indicated earlier, when warranted a summary on data dissemination issues for discussion with the authorities would be prepared by STA prior to the Article IV mission, and included in the SIA accompanying each Article IV staff consultation report.

46. Over the medium-term, data dissemination of the 15 data categories is envisaged for about two-thirds of e-GDDS participants, with the total project cost estimated at about 12–18 full-time (staff) equivalents (FTEs), or about 2½–3 FTEs per year (Table 11). The resource implications are based on the following cost components and staffing considerations:

Table 11.

e-GDDS: Resource Implications

(in units indicated)

article image
Source: IMF staff estimates and projections.

Costs include intitial startup investment of about US$10,000 per participant plus annual fee (about US$3000).

The figure assumes the cost of nearly all e-GDDS participants use of the ODP.
  • During the first year of the project, an initial investment of about US$ 100,000 would be needed to assist in the establishment of the IT infrastructure to monitor e-GDDS participants’ NSDPs. The initial investment would form part of the Fund’s administrative budget for capital projects.

  • About 2½–3 FTEs per year over the medium-term. Staff resources would be allocated to assist e-GDDS participants: (i) collaboration with the data compiling agencies to setup the NSDP/ODP, including pre-populating data files using standardized data definitions, and website or ODP/dashboard design; (ii) a (two-person) mission to work hands-on with each participant;32 and (iii) post-mission follow-up activities, which would include establishing the monitoring process and other technical issues.

  • Somewhat less than 2 FTEs per year to monitor the e-GDDS framework in the post- implementation phase. While the system will largely be automated, staff resources also will be needed to maintain regular contact with the authorities to address monitoring issues (such a metadata updates, technical failures, changes in web and data links).

Issues for Discussion

  • Do IMF Executive Directors agree with staff’s position that the SDDS and SDDS Plus requires no modifications at this time, and that the focus should remain on promoting subscription and adherence, respectively?

  • Do IMF Executive Directors agree with staff’s proposal to enhance the GDDS by aligning the GDDS data categories to the TCIRS?

  • Do IMF Executive Directors agree with staff’s proposal for dissemination thresholds for the e-GDDS, and to monitor progress using these thresholds?

  • Do IMF Executive Directors agree with staff’s proposal that data dissemination issues should be discussed during Article IV consultation missions, when warranted?

  • Do IMF Executive Directors agree that the Tenth Review of the Data Standards Initiatives should be in five years time?

Appendix I. GDDS Participation Date and Recent Developments

Appendix I.

GDDS Participation Date and Recent Developments

article image
article image
article image
Sources: Dissemination Standards Bulletin Board (; various authorities’ national webpages; and IMF Monetary and Capital Markets Department database.

Based on partial information. Proceeds of the capital raised cover government financing needs (excluding debt restructuring); the minimum amount is US$200 million.

Appendix II. How Far Might GDDS Participants be from the SDDS?

1. One way to assess how far GDDS participants may be from the SDDS is through a differential review of per capita income and indicators of institutional strength. Per capita income (GDP per capita) may be a key factor inasmuch as improving statistics and achieving a good track record of disciplined data dissemination requires substantial resources. The strength of institutions (measured, for example, by Polity2 score)33 also may be a key factor in that good governance is important to support transparency and the willingness to commit to and observe international standards. The human development index may be considered a good indicator of capacity development in a country. Against this backdrop, staff conducted a review of how the two groups of countries differ along these dimensions. The sample comprises all SDDS subscribers and 80 GDDS participants for which indicators of these dimensions are available.

2. The relative positions of countries can be visualized as points in the three-dimensional space delineated by the dimensions identified above and shown in Figure 1. A separation between SDDS subscribers (denoted by ☆) and GDDS participants (denoted by ○) is evident in the figure, although a few SDDS outliers are clustered among the GDDS participants (bottom of left-hand-side panel). SDDS countries tend to have higher scores—close to 1—in the three dimensions, suggesting an association of the three dimensions with higher statistical development. While most of the GDDS participants have lower scores along the income dimension, there are many that are clustered with SDDS subscribers along the governance and human development dimensions (left-hand-side panel), suggesting that they may be good candidates to graduate to the SDDS.

Figure 1.