Korea’s statistics are generally of a high quality, and reflect a significant improvement since the last Report on the Observance of Standards and Codes. Korea’s adherence to internationally accepted statistical practices requires cross-cutting recommendations, and the development of a national publication dedicated to the dissemination of data and metadata for general government. For monetary statistics purposes, valuing all marketable financial instruments in the Bank of Korea (BOK) balance sheet and aligning the basis of recording of income transactions to that recommended by international standards is required.


Korea’s statistics are generally of a high quality, and reflect a significant improvement since the last Report on the Observance of Standards and Codes. Korea’s adherence to internationally accepted statistical practices requires cross-cutting recommendations, and the development of a national publication dedicated to the dissemination of data and metadata for general government. For monetary statistics purposes, valuing all marketable financial instruments in the Bank of Korea (BOK) balance sheet and aligning the basis of recording of income transactions to that recommended by international standards is required.

I. Overall Assessment

1. Korea’s statistics are generally of a high quality, and reflect a significant improvement since the last Report on the Observance of Standards and Codes (ROSC) Data Module for Korea in 2001, which itself recognized substantial improvement over the previous few years. This momentum has been reinforced by legal reform, in particular the most recent revisions to the Bank of Korea Act in February 2008, and the Statistics Act in April 2009. The responsibility for macroeconomic statistics is more heavily located in the central bank, the Bank of Korea (BOK), than is the case in many countries, with the BOK holding responsibility for national accounts, producer price indices (PPI), monetary, and balance of payments (BOP) statistics. Statistics Korea (KOSTAT) is responsible for the consumer price index (CPI), reflecting Korea’s inflation targeting policy regime and the undesirability of the central bank to be seen to be responsible for compiling the indicator that it is mandated to target. The compilation of government finance statistics (GFS) is decentralized, with central government falling to the Fiscal Policy Bureau of the Ministry of Strategy and Finance (MOSF), while data on local governments are compiled by the Ministry of Administration and Security (MOPAS) and the Ministry of Education, Science and Technology (MEST).

2. Korea subscribed to the Special Data Dissemination Standard (SDDS) of the International Monetary Fund (IMF) in September 1996 and started posting its metadata on the Dissemination Standards Bulletin Board (DSBB) in March 1998. Korea is in observance of the SDDS, meeting the specifications for coverage, periodicity, timeliness, and the dissemination of advance release calendars. The country uses flexibility options on the timeliness of the general government operations data and the depository corporations survey (DCS). On the other side, it exceeds the SDDS requirements for timeliness and periodicity in some areas. Appendix Table I provides an overview of Korea’s dissemination practices compared to the SDDS.

3. This Data ROSC contains the following main observations. Korean statistics are of a high standard and have improved substantially since the 2001 ROSC Data Module. Plans in train will further establish the strength of the Korean statistical system. In all compiling agencies there is an appropriate legal structure, as well as strong provisions to ensure the integrity of the data. While largely sound, nearly all datasets have some room for improvement. As regards national accounts, improvements could be made in a few statistical techniques. The CPI is sound, although KOSTAT may wish to consider whether to more fully integrate measures relating to owner-occupied housing; also, an unbiased approach in the imputation of out-of-season prices should be adopted. The PPI could be improved by taking the price movements of similar products in the event that a particular price is missing, rather than simply taking the most recent observation. For GFS, until very recently, a key issue was improving accessibility to the data and metadata for measures other than central government. For monetary statistics, the BOK’s foreign currency denominated assets and liabilities are revalued only twice a year, as opposed to revaluation at least monthly, which is common internationally. Finally, BOP statistics could show improvements as regards scope, classification, and basis of recording practices.

4. The compilation and dissemination of statistics are at a key juncture at this time. There are ambitious programs, for instance, to migrate to the Balance of Payments and International Investment Position Manual, Sixth Edition (BPM6) as from 2010, the Government Finance Statistics Manual, 2001 (GFSM 2001) in 2012, and the System of National Accounts, 2008 (2008 SNA) in 2014, and to further enhance the quality of the websites of the compiling agencies. These could help bring Korea towards the cutting edge of statistical achievement. On the other hand, there still remain a range of problems on data accessibility, and progress to the new methodologies could lead to data inconsistencies during the transition if not fully coordinated. It would be helpful to have close collaboration among the compiling agencies, and with users both inside and outside the official sector, to ensure the migration programs can move ahead smoothly and effectively. This could well be a focus for the work of the National Statistics Committee (see below) over the coming years. And, as noted above, Korea still maintains two flexibility options in the SDDS from which it could seek to graduate. The compiling agencies may consider using the opportunity of the prospective 2010 rebasing of national accounts, CPI, and PPI, to effect desired changes smoothly in those datasets.

5. Section IV of this report provides some summary recommendations for addressing these issues. Some of the recommendations in this report can be implemented very quickly; others may take staff and financial resources.

6. In applying the IMF’s Data Quality Assessment Framework (DQAF) July 2003, the remainder of this section presents the mission’s main findings. The findings are presented at the level of the DQAF’s quality dimensions, by agency for the first two dimensions, and across datasets for the remaining four.

7. Pre-requisites of quality and assurances of integrity

  • National accounts, PPI, and monetary and BOP statistics are compiled by the Bank of Korea in accordance with Article 86 of the Bank of Korea Act which states that “The Bank of Korea may, when necessary for the formulation of its monetary and credit policies, collect and compile statistics on money and banking, public finances, prices, wages, production, the balance of payments, and other basic economic statistical series and conduct economic research and for such purposes request any materials or information from government organizations and any juristic or individual person.” Article 42 of the Banking Act states the terms under which commercial and specialized banks are to submit balance-sheet accounts to the BOK and that the BOK may publish the reported data; and the Foreign Exchange Transactions Act and bylaws provide authority for the BOK to collect information on international transactions for the purposes of compiling the BOP. The BOK is an independent agency, protected from interference from government under the provisions of the Bank of Korea Act. The present management has set high store on enhancing this central bank independence. In addition, the Statistics Act specifies all aspects of the production of statistics in Korea. Article 15 describes conditions for the designation of statistics collecting agencies; Article 18 governs conditions for the collection; and Article 27 describes conditions for dissemination of statistics. As a designated agency, the BOK is governed in its statistical work also by other provisions of the Statistics Act. Resources are commensurate for the BOK to carry out its statistical functions, which are organized within a dedicated Economic Statistics Department (ESD). The institution seeks to maintain high levels of transparency in the compilation and dissemination of its statistics, with public explanation of the terms and conditions under which the statistics are collected, compiled, and disseminated, no prior internal government access, and advance notice of major changes. There is heavy emphasis on professionalism among the staff. All BOK staff are bound by the Bank of Korea Act, section 42, which specifies that they should not divulge confidential matters to outsiders. Similar provisions are contained in Article 20 of the Articles of Incorporation of the BOK, as well as in the Rules of Employment, the Code of Conduct, the Code of Ethics and Morals, and the Guidelines of Compilation and Dissemination of Statistics of the BOK and the Foreign Exchange Transactions Act of the MOSF. All staff must take a written oath to commit to follow the Code of Conduct.

  • The Statistics Act establishes KOSTAT as the overarching statistical agency in the country, with powers to designate the conduct of statistical exercises or to designate other agencies to provide specified statistics. Under the Statistics Act a high-level deliberative body, the National Statistics Committee, has been established under the Minister of Strategy and Finance. KOSTAT is mandated to compile the CPI, and has designated the conduct of a consumer price survey (CPS) and a household income and expenditure survey (HIES) in order to be able to fulfill its mandate. KOSTAT is an independent agency, protected by the Statistics Act from outside interference. It has sole power to choose its statistical techniques and processes. A high emphasis is placed on professionalism, with all staff mandated to take 90 hours of training each year, and encouraged to prepare papers and attend international conferences. The Statistics Act requires the Commissioner to secure resources to carry out the agency’s mandate: resources are broadly adequate. KOSTAT staff have to respect the confidentiality of the statistics. There is no pre-release to government. There is also high emphasis on ethics, with newcomers having to attend a course on ethics, and existing staff having to retake the course twice a year.

  • The Government Organization Act and subsequent Presidential Decrees specify the Fiscal Policy Bureau of the Ministry of Strategy and Finance as the agency to compile GFS. Local government data are the responsibility of the MOPAS and local education of the MEST. All ministries meet periodically in order to coordinate statistical activities. As agencies designated by KOSTAT, the staff are bound also by the Statistics Act. As noted above, this ensures the confidentiality of data. Resources are sufficient for the compilation and dissemination of the data, and there are high levels of professionalism. Relevance is ensured by frequent contacts with users. Quality awareness is high, with outside evaluations on key current topics. Efforts are being made also to provide explanations on the consolidated budget balance.

8. The methodological foundations of the statistics are generally sound, and reflect substantial improvement since the 2001 ROSC Data Module. National accounts are closely based on the System of National Accounts, 1993 (1993 SNA), although there could be some relatively minor improvements in scope. The CPI is appropriately based, although farming and fishing households are excluded and the “all items” index does not include estimates from housing owner-occupiers (these are shown separately). The PPI too has an appropriate overall structure in terms of concepts and definitions, as well as in scope, although the commodity classification could be better aligned with international standards. The compilation of GFS still follows the methodology of A Manual on Government Financial Statistics, 1986 (GFSM 1986), but a project to implement the GFSM 2001 framework and methodology is well advanced and is scheduled to result in the publication of fiscal source data in the GFSM 2001 format in 2012. A comprehensive new digital budget and accounting system has been introduced and new government accounting standards have been adopted. Most government units already apply accrual accounting and double bookkeeping (cash based data will continue to be available). Monetary statistics broadly follow the Monetary and Financial Statistics Manual, 2000 (MFSM), although the central bank survey does not identify the monetary base. In addition, the BOK values its financial assets and liabilities at book value and revalues foreign currency assets and liabilities only twice a year. The BOK intends that BOP statistics migrate to the BPM6 framework starting in 2010, making Korea among the first countries to do so. At present Korea follows the Balance of Payments Manual, Fifth Edition (BPM5) framework, although with a number of deviations: The scope of the transactions of enterprises through accounts held abroad and of reinvested earnings is partial; trade credits and short-term loans with affiliated entities are recorded under other investment; and income transactions are recorded on a payment basis. Data with an improved scope of the reinvested earnings data are to be disseminated in early 2010.

9. The accuracy and reliability of Korea’s statistics have improved markedly in recent years. National accounts fulfill most of the requirements for accuracy and reliability, although a better method could be used to adjust original quarterly data to agree with annual estimates, and no specific estimates are included for the non-observed economy. Depreciation should be replaced with consumption of fixed capital for sectors other than general government. The CPI appears accurate and reliable, although there is some scope for improvement in some specific statistical techniques. As regards the PPI, while it fulfils most of the requirements for accuracy and reliability, temporarily missing prices for both regular and seasonal products are imputed by simply repeating the most recent reported observation; also, elementary aggregate indices are currently calculated using the “Carli formula” as the simple arithmetic mean of price relatives, which has been shown to have potential bias. GFS, BOP, and monetary statistics are accurate and reliable.

10. The serviceability of Korean statistics is in many cases excellent, with timeliness in many areas exceeding the requirements of the SDDS, in some cases by significant margins. BOP statistics are particularly strong in this regard, with timeliness better than SDDS requirements across many of the categories. Similarly, the timeliness of the CPI (as well as some labor market indicators) is much faster than required under the SDDS, with dissemination on the first business day in the month after the reference period. On the other hand, Korea maintains two flexibility options. It maintains a flexibility option for the timeliness of general government operations, which are usually disseminated one year after the reference period rather than the mandated two quarters and for the timeliness of the DCS, which is disseminated after six weeks rather than the mandated one month. There are a number of data gaps that users have indicated they would wish to see covered, in particular seasonal adjustment for the CPI.

11. The accessibility of Korea’s data at the moment is somewhat problematic, although there is potential for it to become excellent. Data-compiling agencies have put substantial efforts into developing their websites, so that data can be downloaded quickly and data manipulations performed easily. In the BOK, this is achieved through the Economic Statistics System (ECOS) database, and at KOSTAT through the Korean Statistical Information System (KOSIS). Users interviewed by the mission found the systems excellent. However, these systems encounter a problem that reportedly is affecting computers throughout Korea. These systems are based on access to Internet Explorer, which has a so-called “Active X” component that is intended to increase the speed of access to the site. Upon seeking to access ECOS or KOSIS, users are asked to download a dedicated locally-provided software program. While this software may be cutting-edge in terms of accessing and manipulating data, the “Active X” component on the browser apparently prevents the software crossing the firewalls that protect users, or leads to security warnings before the software can be downloaded. Both the BOK and KOSTAT are aware of this issue, and are working on addressing it. The BOK estimates that the problem will be resolved by 2012, when there will be state-of-the-art browser-free access, while KOSTAT has indicated that it expects that financing to address this will be available by the end of 2010, and it may take until 2013 to resolve. In the meantime, however, international users without direct assistance are restricted to obtaining data and metadata from print publications or by going through commercial vendors. Although the MOSF’s GFS webpage does not use sophisticated browser software, until recently users were obliged to download a program to access the data. Until recently, the accessibility of the GFS lagged considerably behind that of the other datasets, which probably contributed to the concerns raised about GFS in the users’ survey (see section III). Methodological descriptions of the GFS remain dispersed.

12. Section II provides a summary assessment by agency and dataset based on a four-part scale. A summary of the users’ survey is shown in Section III. This is followed by staff recommendations in Section IV. Practices compared to the SDDS are summarized in Appendix I. The authorities’ response to this report and a volume of detailed assessments are presented in separate documents.

II. Assessment by Agency and Dataset

13. An assessment of the quality of six macroeconomic datasets—national accounts, consumer price index, producer price index, government finance, monetary, and balance of payments statistics—was conducted using the DQAF, July 2003. In this section, the results are presented at the level of the DQAF elements and using a four-point rating scale (Table 1). Assessments of the prerequisites of data quality and the assurances of integrity (Dimensions “0” and “1” of the DQAF) are presented in Tables 2ac. For the various datasets, the assessment of methodological soundness, accuracy and reliability, serviceability, and accessibility (Dimensions “2” to “5” of the DQAF) is shown in Tables 3af. The analysis was carried out on the basis of a self-assessment by the compiling agencies, discussions during the mission, and an informal users’ survey.

Table 1.

Data Quality Assessment Framework July 2003—Summary Results

Key to symbols: O = Practice Observed; LO = Practice Largely Observed; LNO = Practice Largely Not Observed; NO = Practice Not Observed; NA = Not Applicable

article image
Practice observed: Current practices generally meet or achieve the objectives of DQAF internationally accepted statistical practices without any significant deficiencies. Practice largely observed: Some departures, but these are not seen as sufficient to raise doubts about the authorities’ ability to observe the DQAF practices. Practice largely not observed: Significant departures and the authorities will need to take significant action to achieve observance. Practice not observed: Most DQAF practices are not met. Not applicable: Used only exceptionally when statistical practices do not apply to a country’s circumstances.
Table 2a.

Assessment of Data Quality—Dimensions 0 and 1—Bank of Korea

article image
Table 2b.

Assessment of Data Quality—Dimensions 0 and 1—Statistics Korea

article image
Table 2c.

Assessment of Data Quality—Dimensions 0 and 1—Ministry of Strategy and Finance

article image
Table 3a.

Assessment of Data Quality—Dimensions 2 to 5—National Accounts

article image
Table 3b.

Assessment of Data Quality—Dimensions 2 to 5—Consumer Price Index

article image
Table 3c.

Assessment of Data Quality—Dimensions 2 to 5—Producer Price Index

article image
Table 3d.

Assessment of Data Quality—Dimensions 2 to 5—Government Finance Statistics

article image
Table 3e.

Assessment of Data Quality—Dimensions 2 to 5—Monetary Statistics

article image
Table 3f.

Assessment of Data Quality—Dimensions 2 to 5—Balance of Payments Statistics

article image

III. Staff’s Recommendations

14. Based on the review of Korea’s statistical practices, discussions with the data-producing agencies, and responses from data users (see Appendix III of the Detailed Assessments volume), the following recommendations are offered. They are designed to increase further Korea’s adherence to internationally accepted statistical practices and would, in the mission’s view, enhance the analytical usefulness of Korea’s statistics. Some additional suggestions are included in the accompanying Detailed Assessments volume.

Cross-cutting recommendations

  • Expedite plans to remove the “Active X” problem affecting the BOK and KOSTAT by which international and other users with firewall protection have difficulty, or are unable, to download software necessary for accessing the data and metadata.

  • Maintain coordination across agencies to ensure data consistency during the migration process towards the new methodologies.

National accounts

  • Use the Denton method to adjust quarterly original estimates to agree with annual benchmark estimates.

  • Replace the existing estimates of depreciation with estimates for consumption of fixed capital for all sectors other than general government. The estimates for consumption of fixed capital should be derived using the Perpetual Inventory Method (PIM) methodology.

  • Periodically undertake an analysis of revisions to movements in quarterly GDP (and its components) in current price and volume terms between preliminary and final estimates from 1999 forward (the first year that seasonally adjusted estimates were published).

  • Investigate the feasibility of developing estimates for the non-observed economy using employment estimates from the Labor Force Survey (conducted by KOSTAT) and employment estimates from other data sources used to compile the national accounts.

  • Aim to remove the statistical discrepancy between the production and expenditure measures of GDP, on an annual basis, by compiling balanced annual supply-and-use tables.

Consumer price index

  • In the context of the 2010 CPI rebase, consult widely with users (especially the BOK) to determine the current principal use of the CPI (e.g., in the measurement of consumption or inflation), define the most appropriate headline measure (e.g., including or excluding owner-occupied dwellings), decide on the preferred treatment of owner-occupied dwellings (imputed rentals or net acquisitions) and define core or underlying inflation. Consider whether to give greater prominence to a broader measure of price changes than the present “all items” index that excludes owner-occupied dwellings.

  • Review the methodology for estimating missing seasonal product prices, with the aim of removing the bias associated with using the simple carry-forward method.

  • As part of the consultations planned for the prospective rebasing program consult on the proposal to change from a five-yearly to a three-yearly rebase cycle by reference to a detailed analysis of the extent of substitution bias associated with the use of the Laspeyres formula. Assess the relative improvement obtained by adopting a three-year cycle compared with that from an annual cycle.

  • Assess the feasibility of including farming and fishing households within the scope of the CPI either by extending the scope of the HIES or incorporating data on expenditure patterns from the Farming and Fishing Household Economic Surveys during the weight calculation exercise.

Producer price index

  • Implement a Korean Central Product Classification, based on the Central Product Classification, for the 2010 PPI rebase in 2013.

  • Introduce the use of more sophisticated techniques for imputing temporarily missing prices for both regular and seasonal products, as described in the PPIM 2004.

  • Continue the compilation and study of the Korean supplementary PPI using the geometric mean formula for the computation of elementary aggregate indices. Implement with the 2010 rebase. To the extent that the component products within an elementary aggregate are heterogeneous, with a resultant disparity of price levels, cluster the prices within such elementary aggregates within narrow price bands, and ensure the units of measurement are comparable.

  • Continue the compilation and study of annual chain Laspeyres PPIs, and implement with the next index rebase to counter the impact of substitution bias.

  • In metadata documentation, explicitly identify areas where there are departures from international standards in the Korean PPI.

Government finance statistics

  • Continue the development of a national (electronic) publication dedicated to the dissemination of data and metadata for general government and its subsectors in Korean and English.

  • Strive to make a detailed sources and methods publication of the Korean GFS compilation available to users in Korean and English.

  • Study the possibilities to publish the general government operations data within the timeliness prescribed by the SDDS, and thus withdraw the SDDS flexibility option.

  • Footnote in all publications of government debt that, in contrast to the functional definition of government in the operations accounts, the central government debt data also include bonds issued by the FESF. Furthermore, involve the BOK and KOSTAT in the discussions on how to make the GFSM 2001 definition of government operational in Korea. Consider classifying the FESF into the accounts of the central government at the time of the GFSM 2001 migration, in line with the precepts of that manual.

Monetary statistics

  • In the future, along with the implementation of the GFSM 2001 methodology, consider recording all FESF flows and stocks in the central government accounts. Ensure consistent treatment of the FESF accounts in the monetary statistics, BOP, IIP, and GFS at all stages.

  • For monetary statistics purposes value all marketable financial instruments in the BOK balance sheet, including monetary gold, on the basis of market prices, and revalue assets and liabilities denominated in foreign currencies at the end of every month.

  • In compiling the ODCs, do not include net claims on the BOK in the consolidation adjustment, but present claims on, and liabilities to, the BOK on the assets and liabilities side, respectively.

  • Take necessary actions to be able to disseminate the DCS within one month after the end of the reference month, in order to withdraw the SDDS flexibility option.

  • Introduce regular cross-checking procedures of positions between the BOK and the ODCs, as well as positions among ODCs.

Balance of payments statistics

  • Expand the transactions coverage by including all transactions on reinvested earnings and foreign-currency transactions of enterprises through accounts held abroad.

  • Improve the classification of direct investment transactions by reclassifying data on trade credits and short-term loans of affiliated enterprises.

  • Align the basis of recording of income transactions to that recommended by the international standards.

Appendix Table I. Practices Compared to the SDDS Coverage, Periodicity, and Timeliness of Data

article image
article image
Note: Periodicity and timeliness: (D) daily or days; (W) weekly or with a lag of no more than one week from the reference data or the closing of the reference week; (M) monthly or with a lag of no more than one month; (Q) quarterly or with a lag of no more than one quarter; (A) annually; and (…) not applicable.Italics indicate encouraged categories.