Mongolia ROSC
I. Data Module

SURVEY What do you think of Reports on the Observance of Standards and Codes?

Contents Executive Summary The Standard Summary of Practices Data Dimension: Coverage, Periodicity, and Timeliness   Real sector   Fiscal sector   Financial sector   External sector   Social and demographic indicators Data Quality, Integrity, and Access   Statistical framework   Source data   Statistical techniques   Serviceability   Accessibility IMF Staff Commentary Box 1. An Approach for the Assessment of Data Quality Appendices  (Use the free Adobe Acrobat Reader to view Tables 1 and 2.) Table 1. Data Categories and Indicators: Comparison of the GDDS Recommendations and National Practices Table 2. Salient Features of the Data Quality Matrix

I.  The Standard

1.  The principal standard against which Mongolia's macroeconomic and socio-demographic data are assessed is the IMF's General Data Dissemination System (GDDS).¹ The primary focus of the GDDS is on presenting a framework for evaluating the need for data improvements, prioritizing such improvements, and providing guidance on data dissemination. The GDDS emphasizes the development of core statistical frameworks and indicators, supplemented by encouraged extensions to core frameworks and indicators. The focus on data quality recognizes the fact that for many countries improvements in data quality are a necessary precursor to enhanced dissemination of data to the public.

2.  The GDDS places emphasis on following sound practices in four dimensions:

• the data dimension (the coverage, periodicity, and timeliness of the data)

• quality of the disseminated data

• integrity of the disseminated data

• access by the public

3.  In addition to the four sectors into which economic data are grouped (real, fiscal, financial, and external), the GDDS includes socio-demographic data. The GDDS provides clear links to the IMF's Special Data Dissemination Standard (SDDS)—a more demanding standard for data dissemination—such that a country could use the GDDS as a step towards subscription to the SDDS. Participation in the GDDS is voluntary. It requires a commitment to use the GDDS as a framework for statistical development; designation of a country coordinator; and preparation of metadata, for posting by the IMF on the Dissemination Standards Bulletin Board² (DSBB).

4.  Mongolia participates in the GDDS and is committed to using the GDDS to develop its statistical system. A GDDS coordinator has been appointed, and Mongolia's metadata, including plans for improvement, have been prepared and are posted on the DSBB. The authorities have been encouraged to make the metadata available domestically.

II.  Summary of Practices

5.  The coverage, periodicity, and timeliness for macroeconomic data in Mongolia are summarized and contrasted with GDDS recommendations in Table 1. As regards the quality of data, including their integrity and accessibility features, an experimental framework identifying six key aspects by which to assess the data is presented in the Box and is applied in Table 2.³

A.  Data Dimension: Coverage, Periodicity, and Timeliness

Real sector

6.  The data generally meet GDDS recommendations on periodicity and timeliness. In the case of national accounts aggregates, annual GDP estimates using the production approach are prepared at current and constant (1995) prices, while in respect of GDP by expenditure, only current price estimates are available. However, estimates for some encouraged components, such as quarterly GDP, have been developed on an experimental basis. Several special surveys have been conducted after 1990 to expand the coverage of the informal sector activities. Nevertheless, further efforts are needed to improve the coverage of small-scale activities, especially services, for the period prior to 1998-1999. A well defined short- and medium-term work program has been established to bring the accounts into closer conformity with the 1993 SNA. The publication in 1999 of a revised series for annual GDP for 1995-98 with 1995 as the base year was an important step in this process.

7.  There is no national consumer price index (CPI) although price indices for the capital city and twenty-one provinces are compiled and published. The capital city CPI is used as the national CPI indicator. This index meets the GDDS recommendation on periodicity and timeliness. Short-term plans call for the development of a national CPI during 2000 as well as updating the CPI basket and weights to 1999. Medium-term plans call for developing a system of producer price indices. This is urgently required to improve the quality of the constant price estimates of GDP.

8.  Labor market indicators of employment and unemployment meet the GDDS recommendations for periodicity and timeliness. However, data on wages and earnings are collected but not disseminated. Current practices for generating statistics have shortcomings. While annual data derived from a household survey are more reliable, the quality of monthly data derived from an establishment (enterprise) survey is questionable as the latter excludes employment in services and non-registered enterprises. To mitigate these shortcomings, medium term plans call for developing labor market statistics through the establishment of quarterly sample surveys.

Fiscal sector

9.  Fiscal data compiled by the Ministry of Finance and Economy (MOFE) and disseminated in the publications of the National Statistical Office (NSO) do not fully meet the GDDS recommendations on coverage, periodicity, and timeliness. Preliminary cumulative monthly data are disseminated to the public within a framework comprising separate data on local, central, and general government operations. However, actual annual data that are publicly disseminated are limited to the general government operations. While the timeliness of these data meets the GDDS recommendation, the classification detail is insufficient. In particular, expenditure on defense/public order and safety cannot be separately identified, and no data on the overall balance and its financing are disseminated. Hence, verification of the consistency of the data on government operations with the banking sector data on claims on government is not possible. A well articulated and time-bound action plan has not yet been developed to improve the overall quality, including public dissemination, of fiscal data.

10.  Data on central government debt are compiled by the MOFE on a monthly basis. However, these data are considered confidential and are not publicly disseminated. In addition, the currency conversion methodology applied to government foreign debt does not comply with the internationally accepted standard of using the end-of-period market exchange rate. Also, the liabilities to the IMF are not included in the public debt data compiled by the MOFE, even though the MOFE is the fiscal agent for Mongolia's transactions with the Fund. Subject to these weaknesses, data on foreign debt outstanding as of June 30 and December 31 are published after their presentation by the MOFE to the Parliament and Cabinet. However, only a total figure is published without disaggregation. Action plans to improve the quality of the data and their dissemination have not been firmly developed. Purchase of a debt management system is being negotiated with international vendors.

Financial sector

11.  The data for the banking system generally meet GDDS recommendations for coverage, periodicity, and timeliness. Specific areas of concern are:

• banks' provisions are included as offsets to specific asset categories, rather than being included in capital accounts (reserves). The unpublished, consolidated balance sheet data show these offsets separately for each asset category, but these details are not available to the public;

• beginning in February 2000, the consolidated balance sheet of the banking sector published by the Bank of Mongolia (BOM) does not include data for seven liquidated or insolvent banks, thus creating a break in the time series;

• the BOM has, in principle, adopted the internationally accepted definition of residency specified in the IMF's Balance of Payments Manual, fifth edition (BPM5). In practice, however, the distinction between foreign and domestic positions is not uniform among all depository corporations. The consolidated balance sheet includes accounts of foreign institutions with those of joint venture enterprises on the asset side. Moreover, the depository corporations use citizenship rather than center of economic interest to determine the residency of individuals;

• credit unions, licensed by the MOFE, are not included in the monetary survey; and

• the BOM does not publish the range of interest rates needed for policy evaluation, but is taking steps to correct this deficiency and expects to expand soon the interest rate measures it disseminates.

External sector

12.  The compilation of balance of payments aggregates generally follows BPM5 guidelines and meets the GDDS recommendations on periodicity and timeliness. Coverage of several components of the balance of payments, however, is deficient largely because of weaknesses in basic source data. Work on a number of fronts is underway to improve source data, although some results, especially those tied to surveys, will be realized only in the medium term. Of special concern are deficiencies in the following areas:

• significant revisions by the General Customs Administration to exports and imports are not distributed by month, resulting in large, unallocated, end-of-year adjustments that adversely affect the accuracy of monthly and quarterly data;

• official transfers, compiled by the Ministry of External Relations (MER), do not separately identify current and capital transfers;

• there are significant gaps in the measurement of foreign direct investment flows and income and related financial account transactions;

• official external debt (public and publicly guaranteed debt) positions, based largely on MOFE records, are not recorded at end-of-period exchange rates which is the international practice. Data compiled by the MOFE do not meet the GDDS recommendations on periodicity and timeliness, and no forward debt service schedules are published;

• the MOFE, Mongolia's fiscal agent for transactions with the IMF, does not include liabilities to the IMF in its compilation of public external debt;

• there are significant gaps in the measurement of private sector external debt;

• the BOM has only limited information on external debt not guaranteed by the government; and

• an international investment position statement is not produced.

The BOM is taking steps to address some of these deficiencies. Especially noteworthy is a survey of enterprises that will substantially improve the measurement of services and foreign direct investment income and related financial account transactions. Also the BOM recently initiated discussions with the MER to identify current and capital transfers. Medium-term plans exist for a Debt Monitoring System to be installed that would help Mongolia meet international reporting requirements.

Social and demographic indicators

13.  The NSO compiles and disseminates indicators on population and poverty that meet GDDS objectives for coverage, periodicity, and timeliness. However, the overall quality, including the reliability and intertemporal comparability, of key poverty indicators needs further improvement. The NSO conducted a Living Standards Measurement Survey (LSMS) in 1998 that established an official poverty line for the seven regions of the country and a census of population and housing in 2000, which provided updated benchmarks for a number of socio-demographic indicators. The census was conducted according to the recommendations and guidelines proposed by the United Nations.

14.  The Ministry of Science, Technology, Education, and Culture (MOSTEC) and the Ministry of Health and Social Welfare (MHSW) compile data that are disseminated by the NSO. Statistics on education are compiled based on the International Standard Classification of Education 1997 (ISCED 97), and morbidity and mortality indicators are based on the International Standard Classification of Diseases 10^th Revision (ICD 10).

B.  Data Quality, Integrity, and Access

Statistical framework

15.  Three agencies are primarily responsible for the production and dissemination of data covered by the GDDS—the NSO, the MOFE, and the BOM. The dissemination of most data is carried out as a public service. All three agencies currently disseminate data through various publications, which are available (with or without charge) by subscription.

16.  The NSO was established as an independent statistical agency under the 1994 Law on Statistics, which was last amended in 1997. The head of the NSO is appointed by the Great State Hural (parliament) for a fixed six-year term, and reports directly to parliament. Under the provisions of the Law, the NSO is responsible for producing and disseminating objective and reliable statistical information, safeguarding the integrity of the statistical system, coordinating official statistical programs to ensure the adoption of international standards and methodology with respect to data collection, processing and dissemination, ensuring the confidentiality of reported data, and providing all users equal access to statistical information. The Law sets out the rights and obligations of the statistical bodies at each level, as well as those of respondents to statistical inquiries, guarantees the confidentiality of reported data and provides for administrative sanctions in cases of a breach of the Law.

17.  The system of official statistics, as defined by the Law, which the NSO is charged with safeguarding and coordinating, is a complex system with many contributors. The NSO, although the central statistical agency, does not, by and large, collect data in the field. Data are collected by the statistical divisions of local and central government administrations, including specialized agencies and ministries responsible for a particular field, such as health or education. Because of the complexity of the operations and the need to coordinate with other agencies, a statistical council has been established under the Law to assist the NSO in maintaining official statistical activities and providing advice concerning uniform implementation of standard methodologies.

18.  The MOFE is responsible for compiling and disseminating data on central government operations and central government debt. The Budget Law of Mongolia of December 21, 1992 requires all district and capital city governments to report to the MOFE on a semi-annual basis on the implementation of their budgets. Similarly, the MOFE is required to prepare semi-annual reports on the implementation of the central government budget. The annual report on the implementation of the central government budget is submitted for the Spring session of the parliament. The Law requires the Treasury Department of the MOFE to consolidate the reports on the implementation of the local and central government budgets. The MOFE provides these consolidated reports to the parliament, government agencies, and other official users. In addition, annual aggregated data from these consolidated reports are disseminated in the NSO's Mongolian Statistical Yearbook.

19.  The BOM is responsible for producing and disseminating data on the financial sector and the external sector. The Central Bank Law of September 3, 1996 grants the BOM the legal right to require the submission of information by banks, financial institutions, and other institutions. The law requires that the BOM ensure the confidentiality of all the information it gathers. Severe penalties are specified for failure to treat data confidentially. In addition to its own charter, the BOM is also governed by the Statistics Law of 1997. That law specifies that the National Statistical Board of Mongolia, acting through the NSO, may delegate data collection authority to other agencies. The BOM has used this procedure to develop a program to improve the compilation of balance of payments data by directly conducting its own surveys of enterprises.

Source data

20.  According to the Law on Statistics, all institutions of the public sector (general government and government enterprises), as well as private enterprises, self-employed persons, households, and individuals must report statistical information to the NSO. The Law sets out the range and type of official statistics to be compiled, and the NSO is charged with the responsibility for developing and ensuring the implementation of the statistical standards, classifications, and methodology for collecting, compiling and disseminating the data. Accordingly, the NSO develops adaptations to local requirements of international standard classification systems, designs questionnaires, and develops and disseminates or reviews and endorses methodologies for data collection and compilation.

21.  Data on government operations, covering both central and local governments, are compiled from administrative records of the MOFE's Treasury Department. Data on expenditures on development projects financed through grants and loans from abroad are recorded in the treasury accounts. Foreign financing data are derived from the records maintained by the MOFE, whereas domestic financing data are imputed as residual.

22.  The monetary aggregates published by the BOM are based on actual accounting records of the BOM and the other depository corporations. Separate balance sheet data for the monetary authorities and for other depository corporations are published by the BOM.

23.  The BOM relies primarily on direct reporting by official agencies, international organizations, and financial sector enterprises (including reports from commercial banks on customer transactions) for the compilation of the balance of payments. These reports were recently supplemented by an enterprise survey to improve the measurement of services and portfolio and direct investment.

Statistical techniques

24.  Weak or incomplete source data mean that estimates have to be made in compiling national accounts aggregates, especially for earlier periods. For monetary statistics, the BOM does not use any sampling or other estimating techniques. Statistical and seasonal adjustments are not used. For the balance of payments, the BOM makes coverage adjustments to the customs data and estimates foreign direct investment transactions from commitments data compiled by the Foreign Investment and Foreign Trade Administration (FIFTA). The FIFTA data, however, do not conform to international standards and have serious deficiencies. Short-term capital flows (e.g., trade credits) are derived from model-based estimates.

Serviceability

25.  The NSO publishes brief descriptions of concepts and methods in each issue of the Monthly Statistical Bulletin and the Mongolian Statistical Yearbook. More detailed descriptions of methodology for selected statistical series are published in the Compendium of Selected Methodology and Classification Descriptions, 1999. Unpublished methodological descriptions for detailed components are available upon request.

26.  Data for the national accounts aggregates and the production index as well as for the CPI are published in sufficient detail to allow the user to make rough cross-checks in order to validate the estimates. The NSO is coordinating an effort to ensure consistency in classification and methodology (where applicable) between data compiled by different agencies and for different purposes.

27.  Government finance data are compiled by the MOFE in a framework that is broadly consistent with the IMF's A Manual on Government Finance Statistics (1986). No documentation of the methodology applied is disseminated by the MOFE. Data on general government operations are disseminated with some classification details that permit limited reasonability checks by the users. However, as no data on the overall balance and its financing are disseminated, cross-checking by the users between the fiscal and the money and banking data is not possible. The reconciliation of the MOFE data on bank financing and the BOM data on the banking sector net claims on government has not been attempted.

28.  The BOM does not provide a publicly available description of methodology for the compilation of monetary statistics and the balance of payments statistics. However, disseminated data broadly conform to the guidelines of the IMF's Monetary and Financial Statistics Manual (MFSM) and BPM5. Data are published in the Monthly Bulletin and the Annual Report and are available on the BOM's web site. Footnotes to the monetary and balance of payments tables published in its Monthly Bulletin and Annual Report, as well as presentation of data in sufficient component detail and time series format, permit users to assess the reasonableness of data.

29.  Balance of payments statistics are based, in part, on the balance sheet data from the monetary survey for some financial account transactions and thus ensures a degree of similarity of instrument and institutional coverage. Data on reserves are checked for consistency with the foreign asset position of the banking system. The BOM uses the BPM5 residency criterion for both the monetary survey and the balance of payments, although in practice, the residency rule is not applied uniformly.

30.  There are plans to strengthen data compilation at the BOM and at agencies that provide data to it. For its part, the BOM is currently reviewing the compilation of monetary statistics in light of the new MFSM and has also embarked on a program to improve the analytical framework for international reserves, exchange rates, and interest rates. In balance of payments, the BOM has undertaken a broad program to improve data compilation through the introduction of a direct survey of enterprises that will improve estimates of services and foreign direct investment transactions. In addition, the General Customs Administration has begun a program to improve the collection of import and export data through the introduction of a system of electronic transmission of data. The MOFE is currently considering the use of a debt monitoring program to eliminate the deficiencies of the current system. Finally, the MER will begin to compile data on grant aid that will distinguish between current and capital transfers.

31.  Social and demographic data are disseminated in sufficient component detail and time series format to permit users to assess reasonableness of data. A post-enumeration survey was conducted to verify the coverage of the 2000 census. However, discrepancies between administrative and survey data have not been reconciled.

Accessibility

32.  The NSO is using an advance release calendar for internal purposes and is planning to publish it as part of the forthcoming catalogue of publications. All data are released simultaneously to all users through press releases, regular publications, such as the Monthly Statistical Bulletin and the Mongolian Statistical Yearbook, and the Internet.

33.  The MOFE does not disseminate statistics on its own. Instead, fiscal data released by the MOFE to the public are published in the NSO's Mongolian Statistical Yearbook and the monthly Statistical Bulletin. Data compiled by the MOFE are widely available to official users prior to their eventual publication.

34.  Data disseminated by the BOM are provided to some government ministries, international organizations and the Parliament ahead of release to the general public and thereafter released simultaneously to other users. Commentaries and analyses are provided for some data series (the monetary survey and the balance of payments) at the time of the release of the data.

III.  IMF Staff Commentary

35.  Mongolia's transition to a market economy necessitated dramatic changes in its statistical system. With extensive external technical assistance, including from the Fund, as well as under the impetus of economic adjustment programs, the authorities have embarked on a program of implementing new, internationally recognized statistical methodologies and establishing a framework for monitoring macroeconomic developments. The NSO, MOFE, and BOM have enhanced their capability to compile and disseminate statistics to meet the needs of both official and private users of statistics. Substantial improvements have been made in all statistical areas.

36.  With the important exception of government finance statistics, including government debt, the statistics currently produced in Mongolia generally meet, and in many instances exceed, the recommendations of the GDDS with respect to periodicity and timeliness of dissemination. However, with respect to their quality and policy value, further improvements are needed, especially in the government finance statistics and, although to a lesser extent, in national accounts and balance of payments statistics. Of particular concern is that data on government operations that are disseminated to the public are limited to above-the-line items and lack component detail. Overall balance and financing data are not disseminated and, as a consequence, the usability of fiscal statistics is diminished. No data are disseminated on government debt, and the quality and coverage of foreign debt data compiled by the MOFE are inadequate. Greater political commitment to fiscal transparency will be required to resolve these issues.

37.  In real sector statistics, further efforts are needed to improve coverage of the informal sector and small-scale activity, particularly in the services sector. Sufficient resources will have to be assured to implement the program for the restructuring, improvement, and expansion of the national statistical system, including the development of the statistical business register, streamlining and integration of the data collection and compilation system, the development of quality control procedures and analytical capability, and the upgrading of the data transmission and dissemination process.

38.  In the external sector, the areas of greatest need for improvement concern the accurate measurement of: (1) merchandise trade by the General Customs Administration, (2) private sector transactions related to services, and foreign direct and portfolio investments (including income), and (3) short-term capital flows.

39.  In financial sector statistics, the greatest challenge appears to be the statistical treatment of the banking sector restructuring. The current practice of excluding financial institutions in distress from the coverage of money and banking statistics results in breaks in continuity of time series and diminishes the analytical value of monetary aggregates.

40.  In the socio-demographic sectors, the greatest need is to address concerns about the overall quality, including the reliability and completeness of coverage, of existing statistics for the data to be of operational use in Mongolia's poverty reduction strategy. This is particularly the case for estimates of key poverty indicators.

41.  Mongolia's decision to participate in the GDDS should provide a new stimulus to the process of transforming its statistical system. As part of their efforts at meeting the recommendations of the GDDS, the authorities have articulated comprehensive plans for improvement in their data compilation and dissemination practices, including prioritization, sequencing, and timing of implementation. The work program to improve the quality of real sector statistics as well as balance of payments statistics is detailed and impressive. Many of the initiatives run through to the medium term and beyond, and the authorities have taken care to indicate that extensive technical assistance is a necessary precondition for undertaking and following through on many aspects of statistical reform. The need is for on-site assistance by experts and provision of necessary software and equipment, and, importantly, training of staff.

Box 1. An Approach for the assessment of Data Quality Statistical infrastructure The institutional and legal foundation for the collection, compilation, and dissemination of macroeconomic statistics. Rationale:  The statistical infrastructure is an important factor underlying the quality of statistics. Conceptual framework The body of features relating to concepts, definitions, coverage, classification, sectorization, and other aspects of macroeconomic statistics that is intended to provide the analytical basis for the compiled statistics. Rationale:  The conceptual framework attempts to define the intended features of the compiled data. A conceptual framework that is in line with international standards facilitates cross-country comparisons. Source data The nature of the source data and the manner in which they are processed. Rationale:  The source data are one determinant of the extent to which compiled data attain their intended features. Statistical techniques The body of statistical estimation, adjustment, and other methods, together with methods of ensuring arithmetic precision, as applied to data collection and processing. Rationale:  Statistical techniques are another determinant of the extent to which compiled data attain their intended features. Serviceability The features relating to requirements of users in terms of general usefulness, consistency, and revisions policy and practices. Rationale:  Data that are in conformity with best practice in other respects also need to meet these tests of usefulness. Accessibility The availability of data and metadata to users. Rationale:  Accessibility includes features of a broad concept of data quality.