Abstract
This Report on the Observance of Standards and Codes (ROSC) on the Data Module on El Salvador reviews summary assessment of data quality. Dissemination of data subject to the IMF’s Special Data Dissemination Standard (SDDS) is in observance of all SDDS specifications, including data coverage, periodicity, timeliness, and the dissemination of advance data release calendars. IMF staff recommends provision of a sufficient budgetary, human resource, and information technology resource allocation to agencies to assure the production of essential source data and significant quality improvements in all areas of macroeconomic statistics.
I. Introduction
1. This data module of the Report on Observance of Standards and Codes (ROSC) contains (1) a brief assessment of El Salvador’s practices with regard to the dissemination of macroeconomic statistics against the Special Data Dissemination Standards (SDDS) of the International Monetary Fund (IMF) (Section II), (2) a summary assessment of the quality of the main macroeconomic statistical datasets, based on the Data Quality Assessment Framework (DQAF) methodology prepared by the IMF’s Statistics Department (Section III), and (3) the main IMF staff recommendations for improving the quality of these data, based on the foregoing assessments (Section IV). The assessments have been prepared on the basis of information provided to IMF staff by the Salvadoran authorities and official data in the public domain.
II. Data Dissemination Practices and the Special Data Dissemination Standard
2. El Salvador subscribed to the SDDS2 in June 1998; El Salvador’s data dissemination practices are therefore assessed against the SDDS. El Salvador is in observance of the SDDS, having complied since October 1999 with the specifications on data coverage, periodicity, timeliness, and the dissemination of advance release calendars. The authorities started disseminating metadata on the Dissemination Standards Bulletin Board (DSBB) in December 1998. Since May 2000, they have been disseminating the Data Template on International Reserves and Foreign Currency Liquidity and have had the template hyperlinked to the DSBB. In August 2000, they established a hyperlink between the national summary data page (NSDP) and the DSBB.
3. El Salvador provides access to the macroeconomic statistics covered by the SDDS in publications and on the websites of the Central Reserve Bank of El Salvador (CRBES) (http://www.bcr.gob.sv), the Ministry of Finance (MF) (http://www.mh.gob.sv), and the Ministry of Economy (ME), of which the DIGESTYC is a dependent agency, (http://www.minec.gob.sv).
Data dimension: coverage, periodicity, and timeliness
4. El Salvador observes SDDS data dimension requirements in all categories, using flexibility options for the periodicity of labor market data. Periodicity and timeliness exceed the Standard for data on general government operations, central government debt, and the international investment position. Timeliness of the Standard is exceeded for consumer prices, the analytical accounts of the central bank, the balance of payments, merchandise trade, and the reserves template.
Access dimension
5. El Salvador meets SDDS requirements for public access to data. Advance release calendars that meet SDDS requirements are disseminated either in hard copy format or on the website of the CRBES and are supplemented by a quarter-ahead presentation of release dates on the IMF’s DSBB. Data are released simultaneously to all interested parties on the CRBES NSDP (http://www.bcr.gob.sv/ingles/estadisticas/nedd_resumen_estadisticas.html).
Integrity dimension
6. The laws, regulations, and decrees under which official statistics are compiled and disseminated in El Salvador are available to the public—in electronic and nonelectronic formats—in Spanish. They provide a normative framework that goes some way toward supporting the integrity of the statistical system.3 Procedures on internal government access to the data prior to public release are disseminated on the DSBB for the data categories to which they apply; according to El Salvador’s SDDS metadata, there is no internal access outside the producing statistical agency prior to public release for any data category covered by the SDDS. Data released by Salvadoran statistical agencies are not accompanied by ministerial commentary.
Quality dimension
7. Summary methodology statements have been posted on the DSBB for most SDDS data categories except labor market and central government operations and central government debt data categories. In addition, methodological information is also disseminated by the CRBES and DIGESTYC. El Salvador’s statistical agencies also disseminate component details and additional data series that make possible crosschecks and checks of reasonableness for all data categories as prescribed by the SDDS.
Monitoring of data and access dimensions
8. Following the IMF Executive Board’s Third Review of the SDDS, IMF staff began in July 2000 monitoring subscribers’ performance under the SDDS with regard to the data dimension and the advance release calendars element of the access dimension. Monitoring is carried out against both release dates in the advance release calendar and metadata.4 During July 2000-July 2003, El Salvador’s dissemination practices were in observance of SDDS requirements.5
III. Summary Assessment of Data Quality
9. This section complements the assessment of performance against the SDDS with a consideration of the quality of the data being disseminated and also focuses on the quality of the data that underpin policy analysis in El Salvador. The IMF’s Statistics Department has developed a tool to provide a structure and a common language to assess data quality.6 The DQAF comprises a generic framework and a set of dataset-specific frameworks. The frameworks cover five dimensions of data quality—integrity, methodological soundness, accuracy and reliability, serviceability, and accessibility—and a set of prerequisites of data quality.
10. The mission conducted an assessment of six macroeconomic datasets (national accounts, the CPI, the IPRI,7 balance of payments statistics, government finance statistics, and monetary statistics). The information resulting from the application of this framework to the Salvadoran statistical system is presented below, following the structure of the DQAF. Conclusions are also presented in the form of standardized summary tables in which the assessment of data practices is made on a qualitative basis, using a four-part scale (Tables 1–6 of the Appendix).
11. Statistical development in El Salvador was strongly motivated by the country’s effort to subscribe to the SDDS and come into observance, a process that reached virtual completion in mid-2000. Since then, the authorities have accorded statistical development a relatively lower national priority in terms of policy motivation and appropriate budgetary allocations. A Coordinating Committee for the National Statistics Service (NSS), which was created by presidential decree in July 2000 to prepare a national statistical plan and coordinate statistical activity among statistical agencies, has not met for a long period.
Prerequisites of quality
This category of the DQAF identifies conditions, within the agency in charge of producing statistics, that have an impact on data quality. The elements within the category refer to the legal and institutional environment, resources, and quality awareness.
12. The macroeconomic statistical system of El Salvador is in practice centralized in three agencies, the CRBES, DIGESTYC, and MF. The legal environment,8 which relies on an almost 50-year old Law on the National Statistics Service, is not well suited to a modern statistical system. While this Law establishes a NSS—defined as all public and private agencies that produce statistics—and requires the compilation of most macroeconomic statistics that are now considered to be important for macroeconomic policy analysis, it makes no provision for a specific statistical planning process or the allocation of responsibilities among component agencies of the NSS. Moreover, it applies several useful provisions—such as the obligation to report data and guarantees of confidentiality of data reported—only to the DIGESTYC rather than to all component agencies of the NSS. The Law empowers the DIGESTYC to develop technical standards for the component agencies of the NSS and to collect, compile, and disseminate statistical information, but leaves the agency (established as a dependent General Directorate of the ME) neither the autonomy nor the resources to fulfill these roles.9
13. The Law assigns to the DIGESTYC responsibility for construction, transportation, agricultural, industrial, trade, and cost of living statistics, as well as censuses and a range of socio-demographic statistics. It requires the DIGESTYC to establish statistical standards and methodologies that all component agencies of the NSS will apply and to work toward interagency compatibility of statistics. The Law specifies minimal qualifications for the positions of Director General, Deputy Director General, and General Secretary of the DIGESTYC but makes no provision for their appointment—in practice, they are appointed by the Minister of Economy. The Law requires timely publication of a specified set of statistical publications and the provision of data to users who request them. Since the DIGESTYC is responsible for the CPI, mechanisms for collaboration with other data-producing agencies have not been developed in this area. However, the CRBES also produces an unpublished monthly CPI; elimination of this supplementary CPI would reduce respondent burden and avoid duplication of effort. In other areas also, such as national accounts, establishment of the various committees called for in the legal framework could facilitate closer coordination of statistical initiatives. The provision of data to the DIGESTYC is required by law, although the sanctions for noncompliance, which are not inflation-proofed, are now ineffective. The DIGESTYC also works to assure good faith and awareness of the importance of reporting on the part of data providers.
14. The DIGESTYC has sufficient human and information technology resources for compiling the CPI, although some of the computers and software need to be updated. It is somewhat short of financial resources for this purpose and significantly short of all types of resources for meeting its other statistical obligations. In particular, the major economic and demographic censuses (the former providing essential source data for the national accounts and price indices) have severely languished for lack of financial resources, and this factor has adversely affected the quality of the national accounts in particular. The level of the staff’s academic training is deemed adequate. Measures are in place to make best use of advancing technology, and regular tests of processes are conducted to assure the efficient use of resources. The DIGESTYC demonstrates its quality awareness in the area of the CPI through regular training, quality testing, and double data entry; however, it does not use a separate advisory body on data quality, has not conducted user surveys, and does not have an institutional quality awareness program.
15. The CRBES, which is established as a public autonomous institution of a technical nature, is in practice responsible for compiling and disseminating balance of payments, industrial price, monetary, and national accounts statistics. It is also authorized to compile and disseminate public finance statistics, which are also compiled by two units of the MF. In terms of the legal environment, the Law on the Central Reserve Bank of El Salvador requires the CRBES to compile and disseminate in a timely manner balance of payments, monetary, and exchange statistics, as well as other macroeconomic statistics that the CRBES’s Council may deem necessary, and provides for the Council to determine the nature, content, and frequency of these statistics. The Superintendency of the Financial System (SFS), which is established as an institution integrated into the CRBES but with administrative, budgetary, and functional autonomy, is required by the Law on the Superintendency of the Financial System of El Salvador to disseminate statistical data for all entities under its supervision. It is also empowered to share with the CRBES detailed information on entities under its supervision. Through largely less formal arrangements (primarily the establishment of committees and working groups), the CRBES has ensured sufficient information interchange with other public agencies and private entities to meet day-to-day requirements, although the establishment of more formal collaboration agreements, particularly with the MF, would greatly facilitate longer-term improvements in data consistency and reliability, as well as efficient use of resources.
16. The Law on the Central Reserve Bank of El Salvador mandates that all government agencies and financial institutions report to the CRBES in a timely manner any statistics that the CRBES requires to fulfill its functions and provides not insignificant sanctions—which, however, are not inflation-proofed—for noncompliance. In contrast, the CRBES does not have the legal power to require reporting on the part of municipalities and the nonfinancial private sector, a limitation that impedes the collection of some of the information required for the balance of payments, the IPRI, government finance statistics, and national accounts, although the CRBES provides significant incentives for voluntary reporting. The CRBES cannot rely on the legal provisions of the Law on the National Statistics Service to require the reporting of information, since, as already noted, the relevant provisions of that law apply only to the DIGESTYC.10 The Law on the Superintendency of the Financial System of El Salvador provides strong support for the obligatory provision of information to the SFS on the part of all entities subject to supervision by the Superintendency and sanctions for noncompliance. The Banking Law also requires timely release, on the part of the banks, to the CRBES and the SFS, of all information that these agencies require to fulfill their mandates and provides adequate sanctions for noncompliance. An Interagency Financial System Committee (made up of the President of the CRBES and the Superintendents) meets at least once a month to coordinate information exchange among other things. There is continuous communication between the CRBES and reporting entities to try to alleviate reporting problems, and the CRBES provides adequate support for the preparation of information returns. Legal provisions guarantee the confidentiality of undisclosed information provided to the CRBES and the SFS by reporting entities, and the two agencies have taken appropriate steps to assure this confidentiality.
17. In terms of information technology resources, the CRBES has made efforts to upgrade the computer systems used in the compilation and analysis of statistics. However, there is as yet no institutionalized functioning database for balance of payments, government finance, and national accounts statistics. The transitions to the new Monetary and Financial Statistics Manual (MFSM), from the 1986 to the 2001 edition of the Government Finance Statistics Manual, and from the 1968 to the 1993 System of National Accounts, together with the completion of work on the institutional database, will require additional human and financial resources, including for training in some areas. The CRBES is aware of the need for efficient use of resources and has achieved enhancements in efficiency by use of technological advances and periodic assessments of processes, accompanied by reductions in staffing. Such reductions may have gone too far in some areas; the number of staff assigned to the compilation of monetary statistics, for example, has undergone a significant reduction. The CRBES is demonstrating quality awareness through internal user surveys and a current internal evaluation of strengths, opportunities for improvement, weaknesses, and risks and has taken concrete steps to improve the supply of source data for the balance of payments, monetary statistics, and the national accounts. However, the CRBES does not conduct regular external user surveys for the purpose of obtaining comments on data quality, and statistical processes are not externally audited. In addition, an institution-wide quality awareness program introduced some time ago has lost momentum.
18. The MF has an adequate legal environment, in the Law on Public Financial Administration and its associated Regulations, for the compilation of public sector accounts and public sector debt statistics but not for the compilation and dissemination of government finance statistics according to international standards. The Law prescribes that the MF and the CRBES will collaborate closely in monetary and financial programming.11 The Law establishes an Integrated Financial Administration System (IFAS) and a Government Accounting Subsystem. It requires all public sector entities to report their financial balance sheets and their transactions and other flows to the General Directorate of Government Accounting (GDGA) and endows the GDGA General Directorate with the power to determine general accounting rules for the public sector and the format and content of all financial statements. It also requires the General Directorate to consolidate the data on the public sector for subsectors and the sector as a whole and to facilitate the integration of these data into the national accounts. It requires the MF to maintain the official register of public debt.
19. The MF is responsible under the Political Constitution and the Law to keep the Legislative Assembly informed concerning budget execution. Although it is not required to disseminate government finance statistics to the public, in fact two units of the MF—the Advisory Macroeconomic and Fiscal Unit (AMFU) and the GDGA—compile government finance statistics. The former reports government finance statistics to WHD and the latter reports them for publication in the Government Finance Statistics Yearbook (GFSY). The MF maintains a close ongoing relationship with the CRBES and other public agencies in terms of sharing statistical information. However, more formal coordination with the CRBES would facilitate reconciliation of financing data and longer-term improvements in public sector statistics, including the elimination of duplication in data gathering and compilation. In addition, closer internal coordination between the GDGA and the AMFU would ensure improved verification of published data.
20. The MF has allocated sufficient human, financial, and information technology resources for the compilation of government finance and public debt statistics. Greater efficiency of resource use is slowly being achieved by increasing the use of automated processes. The MF is fully aware of the importance of the quality of statistics, which is being promoted mainly through the establishment of common accounting standards for all budgetary agencies and the automation of the compilation and verification of statistics. The MF does not conduct user surveys with a view to obtaining comments on quality.
Integrity
Integrity identifies features that support firm adherence to objectivity in the collection, compilation, and dissemination of statistics so as to maintain users’ confidence. Elements refer to the professionalism and ethical standards that should guide policies and practices, which should be reinforced by their transparency.
21. Professionalism is a guiding principle of the DIGESTYC. The selection of data sources and methods is dictated purely by statistical considerations. However, the DIGESTYC is not authorized to make technical comments to communications media that have made errors in interpreting the data except through the Minister of Economy, and in practice the procedure is too cumbersome for it to do so. The DIGESTYC’s statistical policies and practices are transparent as the agency does not allow any national authority access to statistics before their release, announces substantial changes in the methodology, statistical techniques, or source data used in advance, and identifies itself as the statistics-producing agency. The ME informs the public of the terms and conditions under which data are compiled, while not disseminating internal regulations. As regards ethical standards, DIGESTYC staff members are subject to the provisions of the Law on the National Statistics Service and internal regulations; the latter need to be better articulated.
22. The CRBES is committed to maintaining the highest standards of professionalism. The legal framework governing the CRBES is adequate for ensuring the autonomy of staff members assigned to the compilation of statistics and professional competence is a key requirement in the hiring of employees and in staff promotions. The selection of data sources is dictated purely by statistical considerations. The CRBES has the right to clarify erroneous interpretations of statistics on the part of the communications media and exercises that right. In general, CRBES statistical policies and practices are transparent as the CRBES informs the public of the terms and conditions under which the statistics are compiled, does not give any national authority access to the statistics before their release, and identifies itself as the statistics-producing agency. However, the CRBES does not publish its Law on its own website12—the Law has to be sought on the website of the SFS—and does not systematically inform the public of the availability of weekly monetary statistics or make them available to users unless they request them. Nor does the CRBES announce any substantial change in the methodology,13 statistical techniques, or source data used in advance but only at the time the new data are disseminated—and in the case of government finance statistics not even then. As regards ethical standards, CRBES staff members are subject to the provisions of the Law on the Central Reserve Bank of El Salvador. In addition, the CRBES has drawn up Labor Regulations, a copy of which is given to all new employees and which is posted on the CRBES’s Intranet site. These regulations explain staff members’ rights and obligations and contain clear sanctions for nonobservance of obligations.
23. In practice, the MF maintains a high level of professionalism, even though technical independence in the compilation of statistics is not backed by laws or specific safeguards. Government finance and public debt statistics are compiled on the basis of accounting records and administrative reports, drawn up in accordance with the relevant charts of accounts and budgetary standards, although data published by the CRBES are not drawn directly from accounting sources. The selection of sources and methods appears free of any political influence. In cases of improper interpretation of the data on the part of the communications media, the MF issues clarifications. The MF’s statistical policies and practices are transparent to the extent that the MF informs the public of the terms and conditions under which the statistics are compiled and does not allow any national authority access to the statistics before their release by the CRBES. Substantial changes in the methodology, statistical techniques, and data sources used are not disseminated by the MF. MF staff members are subject to the provisions of the Law on Public Financial Administration and the Regulations issued under it; there is no other set of ethical standards for the MF, other than the Civil Service Law.
Methodological soundness
Methodological soundness refers to the application of international standards, guidelines, and agreed practices. Application of such standards, which are specific to the dataset, is indicative of the soundness of the data and fosters international comparability. Elements refer to the basic building blocks of concepts and definitions, scope, classification and sectorization, and basis for recording.
24. The concepts and definitions of the national accounts follow the System of National Accountsm, 1968 (1968 SNA), but international assistance is being sought to prepare a migration project to the System of National Accounts, 1993 (1993 SNA). Scope is narrower than the 1968 SNA, since the compilation of the institutional sector accounts, from production to accumulation, was partially interrupted in 1995 and completely abandoned in 1998. Consumption of fixed capital is not estimated. However, some important features of the 1993 SNA have been incorporated, e.g., supply-and-use tables with data valued at current, previous year, and reference period prices. Classifications, by functions for government final consumption expenditure and by purposes for household consumption, are not applied; however, the latter follows the product classification of the supply-and-use table. The national industrial activity classification is based on revision 2 of the International Standard Industrial Classification (ISIC). In general, the basis for recording follows internationally accepted standards. Basic or producer prices are used to value market output and purchase prices are used for intermediate consumption. However, time series at current prices are valued in U.S. dollars, whereas time series at constant prices are valued in colones. Transactions are generally recorded on an accrual basis.
25. Both the CPI and the IPRI meet international standards with regard to concepts and definitions, practices, and standards. Data for estimating base period weights, as well as price data obtained from monthly price surveys, are collected and processed in accordance with international recommendations for both series. The scope of the CPI includes all types of urban households, all monetary consumption expenditure, goods produced for own consumption, and imputed rent for owner-occupied housing. Although the IPRI only covers the manufacturing sector of the Salvadoran economy, that sector accounts for more than 95 percent of the core components of mining, manufacturing, energy, and water production normally covered by this type of index. Both indices have outdated classification systems. Rather than using the current international standard Classification of Individual Consumption by Purpose (COICOP), the CPI employs a system that is more than two generations outdated. The IPRI uses revision 2, rather than revision 3, of ISIC. The classification system used for products for the IPRI is also seriously outdated. The basis of recording for both series is consistent with international best practices. All expenditure, production/sales, and price data are reported in properly specified market, and/or transaction values, on an accrual basis.
26. Balance of payments statistics are compiled in conformity with the methodological standards, concepts, and definitions described in the fifth edition of the Balance of Payments Manual (BPM5). As part of the effort to improve methodological soundness, the CRBES has revised compilation procedures for trade and insurance. On scope, considerable progress has been made in the coverage of some items in the current account, such as travel and other services, while data available on private external debt have been added to the financial account. However, recommendations yet to be implemented include the recording of assets held abroad by residents and the recording of some banking transactions on the basis of residency criteria. Transactions of the general government do not follow the sectorization recommended by BPM5. Interest, dividends, and external debt transactions are still recorded on a cash basis.
27. The data template on international reserves and foreign currency liquidity is generally compiled in accordance with the Operational Guidelines for the Data Template on International Reserves and Foreign Currency Liquidity. However, some significant classification anomalies appear. Section II on predetermined short-term net drains includes items not related to Section II, such as revenues from future placements of bonds in the international markets and future loan disbursements from international financial organizations. In addition, liquidity requirements on foreign currency deposits held with the CRBES by resident banks do not fully follow the concepts and definitions of the Operational Guidelines. The CRBES includes the full amount of the liquidity requirement in gross reserves and does not register the contingent net drain, as specified in Section III of the data template.
28. The concepts and definitions used for the compilation and dissemination of government finance statistics are based on the recommendations of A Manual on Government Finance Statistics 1986 (GFSM 1986). Although the CRBES authorities have not prepared a migration plan to move to the Government Finance Statistics Manual 2001 (GFSM 2001), the MF authorities, particularly the GDGA, have initiated an assessment of the needed changes for implementation of the new framework. On scope, government finance statistics disseminated cover all units of the nonfinancial public sector, except that some decentralized institutions, such as the Salvadoran Institute of Municipal Development, are not included and mutual funds are wrongly included. There is no functional classification of expenditures or financing data by type of debt holder and type of financial instrument. The public debt data disseminated by the CRBES has complete coverage, but those compiled by the MF do not include all short-term liabilities. In terms of classification/sectorization, consolidated central government data disseminated do not include all extrabudgetary units and social security institutions, which are included in the rest of general government. The classification of government finance statistics is broadly in accordance with the recommendations of the GFSM 1986, with some exceptions, such as pension payments and redemptions of recognition bonds incorrectly classified as financing items, inflows from privatization not being treated consistently, and guaranteed debt being recorded as a debt of the consolidated central government. The basis for recording data is mixed. The CRBES data have revenues on a cash basis and expenditures on a check issued basis (a close proxy for a cash basis). All financing data are recorded on a cash basis. On the other hand, data provided by the MF to the IMF for publication have revenues on a cash basis and expenditures on a payment order basis (a close proxy for accrual recording).
29. Although the analytical framework for monetary statistics reflects concepts and definitions that broadly follow the MFSM, the full adoption of the MFSM methodology has not been completed. The scope of the depository corporations survey is currently limited to the CRBES, commercial banks, and financial companies, and excludes other nonbank deposit-taking institutions, and—of greater quantitative importance—money market funds. The classification and sectorization of financial instruments, which is generally consistent with MFSM recommendations, deviates in several ways: financial derivatives, other equity, and insurance technical reserves are not separately identified; and deposits, loans, and other accounts receivable/payable are not fully sectorized. The basis for recording flows and stocks is largely consistent with the MFSM, although some asset and liability positions are valued at acquisition cost rather than market prices. Accrued interest is not included in the outstanding value of financial instruments but in other assets/other liabilities.
Accuracy and reliability
Accuracy and reliability identifies features that contribute to the goal that data portray reality. Elements refer to identified features of the source data, statistical techniques, and supporting assessments and validation.
30. Source data for the national accounts are scarce. Basic data for the accounts have been eroded by the weakening of the DIGESTYC. Its last agriculture census was conducted in 1971 and its annual economic surveys, a vital data source for the national accounts, were interrupted in 1999. Although the CRBES conducts its own surveys, they only partially replace the DIGESTYC surveys, since they cover only medium and large firms and investigate fewer variables than those needed for the national accounts. Annual source data and short-term indicators are especially limited for service activities. Data collection from local authorities has improved. In terms of statistical techniques, annual macroeconomic aggregates are estimated in the framework of the supply-and-use table, which permits the detection and elimination of discrepancies in order to achieve consistency. However, excessive use is made of fixed coefficients more than ten years old. Also, the adjustments to account for the non-observed economy are based on old ratios. The benchmarking technique applied to combine annual and quarterly GDP data is not appropriate, but a good procedure is used to seasonally adjust GDP estimates. Assessment and validation of source data are only partial, and sampling errors are not estimated in some surveys, while assessment and validation of intermediate data and statistical outputs are reasonably comprehensive using the commodity flow method applied in a systematic way in the framework of the supply-and-use table. Accuracy of estimates generated by surveys is not always assessed, as survey errors are not available for some of them. Analysis of revisions is carried out systematically, following the cycle of national accounts estimates, but revision studies are not undertaken regularly.
31. Although monthly price data for both the CPI and the IPRI are adequate and collected in a timely and efficient manner, there are serious deficiencies in source data for constructing sampling frames for essential surveys upon which weight calculations depend. The current CPI weights are based on the 1990/91 National Survey of Household Income and Expenditure (NSHIE). The construction of a household sampling frame for a new NSHIE depends on a new population census (last taken in 1992) and/or updated cadastre maps. The current IPRI weights are based on the 1993 Economic Census. A new economic census requires an updated business register. An updated business register and cadastre maps have been completed for only part of the country. The statistical techniques employed for the compilation of the CPI are generally in accordance with internationally accepted standards. However, the lists of market basket items and the weights for the CPI are in need of updating to reflect current expenditure patterns. The weights for the IPRI are only narrowly within the recommended ten-year limit and were not updated to the index reference period. In addition, there are technical problems with the application of the index formula. Also, the IPRI provides indices by product aggregation only, and not by economic activity. Assessment and validation of source and intermediate data are undertaken for both series. No revision studies have been conducted for the IPRI. Although no recent revision studies have been conducted for the CPI, estimates of variance for expenditure data from the 1990–91 NSHIE are available for use in designing the new NSHIE and the new CPI market basket.
32. The mix of primary and secondary data sources is broadly sufficient to compile major items of the balance of payments statistics. Major administrative source data are available on a timely basis, and quarterly surveys are based on a well-established business registry. However, there are limited data sources to compile holdings of external assets by residents, and the narrow coverage of major companies in the communication, construction, and transportation sectors contributes to weaknesses in the services account. Some items in the balance of payments are estimated using sound statistical techniques, such as travel, which is compiled through sample techniques conducted twice a year. However there are no estimates for nonrecorded trade. Data are assessed and validated against other data, and statistical discrepancies are investigated. Revision studies are seldom undertaken and do not follow a regular schedule.
33. Source data for the compilation of government finance statistics disseminated by the CRBES are collected from a comprehensive set of public units. On the other hand, the source for data reported by the MF to the IMF for publication is the IFAS, which records on an accrual basis and covers detailed financial information of all budgetary units, decentralized units, and some public enterprises. The source data for disseminated government finance statistics follow the budgetary classifications and for financing data are mostly in accordance with the GFSM 1986 definitions and classifications. The compilers are not fully aware of the differences between source data and government finance statistics concepts. The IFAS has no provision for automatic derivation of government finance statistics. Accepted statistical techniques, including extrapolations for missing data and data consolidation, are applied to the set of preliminary data. For local government, data on 262 municipalities are obtained from central government transfers. These transfers are allocated taking into account a specific expenditure classification prepared by the CRBES. No surveys are used to obtain local government data. Assessment and validation of source data are performed through internal and external auditing processes; the Court of Accounts provides external auditing services. However, no systematic procedures are in place to assess the accuracy of disseminated government finance statistics. In terms of assessment and validation of intermediate data and statistical outputs, data sent to the IMF for publication are not validated against data published by the CRBES. No revision studies are undertaken.
34. The data sources for monetary statistics are generally adequate. The plan of accounts for commercial banks follows international accounting standards, and its limited instrument and sectoral disaggregation is complemented with additional information sought by the CRBES. However, the distinction between resident and nonresident accounts in the plan is not explained and not applied to all instruments. The accounts of the CRBES provide sufficient detail to approximate the recommended classification of accounts of the MFSM. The use of statistical techniques is adequate, being confined to automated checks in the CRBES monetary database to identify processing errors. The assessment and validation of source data, which consists of checking the reported balance sheet data against secondary sources, is also adequate. Revisions, which rarely occur, are always investigated, although compilers of monetary statistics do not routinely analyze or compare preliminary and final data and do not undertake studies of revisions because of their insignificance.
Serviceability
Serviceability focuses on practical aspects of how well a dataset meets users’ needs. Elements refer to the extent to which data are relevant, produced and disseminated in a timely fashion with appropriate periodicity, are consistent internally and with other datasets, and follow a predictable revisions policy.
35. The relevance of national accounts for users is not monitored in a systematic way. Nevertheless, frequent contacts are made with the main users and their specific demands are attended to when possible. Timeliness and periodicity follow the dissemination standards of the SDDS. Due to the supply-and-use framework in which they are estimated, aggregates are internally consistent. Also time series are consistent over time. However, statistics at current prices are not reconcilable over a reasonable period of time owing to the different currencies used. The national accounts are fully consistent with balance of payments statistics, but no systematic checks are performed to reconcile them with government finance statistics. In terms of revision policy and practice, the revision cycle follows a regular and well-established two-year schedule. Preliminary data are clearly identified, and the revised data are disseminated with the same level of detail as previously published. Explanations of revisions are disseminated to the public in insufficient detail.
36. For both the CPI and IPRI there is little active monitoring of relevance with regard to user needs in the form of user advisory groups and user surveys. However, public requests for unpublished data provide the DIGESTYC and the CRBES with information on users’ needs. Personnel involved in the compilation of both indices participate in international seminars and meetings. Both price indices meet SDDS standards with regard to timeliness and periodicity. Both indices are internally consistent and reconcilable over time and with other comparable statistical series. The IPRI follows a regular, well-established, and transparent schedule with regard to revisions of the monthly price series; no preliminary data are published for the CPI. Both the 1993 revision of the CPI and the 1998 initial publication of the IPRI are well documented, but no recent revision studies have been conducted for either series.
37. The CRBES does not conduct surveys targeted to external users to monitor the relevance and practical usefulness of balance of payments statistics. However, public requests for unpublished data provide the CRBES with information on users’ needs. Monthly template data on international reserves and foreign currency liquidity, as well as quarterly balance of payments and international investment position statistics, meet the timeliness and periodicity requirements of the SDDS, international investment position data exceeding the periodicity requirement. Statistics are consistent over time and among various external sector accounts. There is full consistency with national accounts and monetary statistics, but reconciliations with government finance statistics cannot be fully effected. Revisions follow a regular, well-established, and transparent 24-month schedule, and preliminary data are clearly identified, but revision studies and analysis are not disseminated.
38. While the government finance statistics timeframe is aligned with the budget and accounting cycle, the coverage and level of detail of government finance statistics could be improved to enable users to better assess government social and fiscal policies. Although the methodological basis and classification structure appear adequate for the analytical needs of users, the relevance and practical utility of government finance statistics are not monitored on a regular basis, and no user surveys are conducted. Timeliness and periodicity of disseminated data follow, and in some cases exceed, SDDS recommendations. Consistency between the overall deficit/surplus and financing data is not achieved because revenue, expenditures, and financing are recorded using different bases and the government finance statistics sectorization of nonfinancial public sector institutions is different from that used in the monetary statistics. Government finance statistics use monetary statistics, in place of treasury records, as the source for data on domestic financing; thus no issue of reconciliation arises. Government finance statistics are not reconciled with the balance of payments and national accounts. Data on domestic debt produced by the MF are reconciled with CRBES data. In terms of revision policy and practice, preliminary data become final the second time they are published. Hardcopy publications do not distinguish between preliminary and final data. However, the CRBES website makes the distinction.
39. Formal processes of consultation with users to obtain feedback on the relevance of monetary statistics could be strengthened, for example, by conducting user surveys and periodic meetings with external users. The timeliness and periodicity of published data meet SDDS requirements, and the data are consistent internally and over time. Consistency checks between monetary statistics, government finance statistics, and balance of payments statistics are conducted regularly. However, the definition of the institutions of the general government is not fully consistent between monetary and government finance statistics. The revisions policy follows a regular and transparent schedule. Analyses of revisions are not made public because revisions are generally not important. Significant revisions are footnoted in statistical publications.
Accessibility
Accessibility deals with the availability of information to users. Elements refer to the extent to which data and metadata are clear and easily available and assistance to the users is adequate to help them find and use the data.
40. National accounts annual and quarterly data are easily accessible on the CRBES website and publications. The estimates are presented in a clear manner in charts and tables. Published quarterly series are seasonally adjusted. A press release is followed by a more comprehensive dissemination in the CRBES’s Quarterly Review. Quarterly national accounts are released on a preannounced schedule, but annual data are not. Estimates are made available to all users simultaneously. Unpublished (but nonconfidential) subaggregates are made available upon request. Documentation on concepts, scope, classifications, basis of recording, and statistical techniques is available at the CRBES’s Communications Department. However the metadata for annual national accounts are not sufficiently detailed. The name and contact information for a contact person is not publicized, but assistance to users is coordinated through the Communications Department. A list of available publications is available on the CRBES website.
41. Data accessibility is generally good for the CPI and the IPRI. Index publications and the CRBES website are well designed and provide sufficient information on the indices. Indices are released simultaneously to all users on a preannounced schedule, and unpublished, nonconfidential data are available upon request from the CRBES and/or the DIGESTYC. Although adequate documentation and descriptive material are available for both index series, metadata accessibility to users is somewhat limited since these metadata are not available on the CRBES website, where both indices are disseminated. Contact persons are available for both index series only in the sense that the organizational heads of the offices where the series are compiled are listed either on the CRBES website or in DIGESTYC publications. No telephone numbers or e-mail addresses are given for either series, and all assistance to users at the CRBES is coordinated through its Department of Communications. Catalogs of statistical products are available for both series.
42. Data accessibility is adequate for balance of payments statistics. Data are disseminated both as hardcopy format and on the CRBES webpage. Detailed quarterly balance of payments and international investment position statistics are released simultaneously to all interested parties through the CRBES website and in official hardcopy publications. There is a preannounced schedule of publication release dates. Regarding metadata accessibility, a brief technical note on balance of payments statistics is disseminated in the Quarterly Bulletin, although no other metadata are available to users. A summary of methodology is included in the IMF’s Balance of Payments Statistics Yearbook. The MF, as a producer of public external debt statistics, does not disseminate statistics or methodological notes. Assistance to users is adequate in the CRBES, although the name and contact information for a contact person responsible exclusively for balance of payments is not disseminated.
43. A wide range of government finance statistics is accessible through quarterly and monthly publications and through the CRBES website. However, there are some deficiencies in format.14 The government finance statistics disseminated in the monthly publication follow a preannounced calendar, but quarterly and annual government finance statistics do not. The date of release of government finance statistics aggregates is announced in the preceding government finance statistics publication. Data are released simultaneously to all users by posting on the CRBES website. Unpublished, nonconfidential, disaggregated data are made available upon request, but the availability of unpublished data and the terms and conditions on which they are made available are not publicized. A brief summary of government finance statistics metadata, except for public debt, is published by the CRBES. Although a more complete internal document describes the concepts, scope, classifications, basis of recording, data sources, and statistical methods used, this document requires significant improvements. For example, there are no bridge tables showing the links between source data and government finance statistics. Information on methodology can be obtained on request from the CRBES, but its availability is not publicized. In terms of assistance to users, the name and contact information for a contact person is not published. The CRBES provides support to users by informal communications. The CRBES website presents a list of available publications.
44. The presentation of monetary statistics in CRBES publications generally meets user needs for data accessibility. Monthly data are published electronically on the CRBES website and in hardcopy publication. Nonconfidential unpublished data are made available upon request, although the availability and the terms and conditions of this service are not made public. The dissemination formats on the website and in publications are user friendly. There is a preannounced schedule of publication release dates. Comprehensive metadata to complement those on the DSBB are not available to meet the needs of more specialized users. In order to assist users, a general electronic mailbox is available on the CRBES website, but without the name or contact information for a contact person on monetary statistics. The set of statistical publications of the CRBES, which is free of charge, is available on its website.
El Salvador: Data Quality Assessment Framework—Summary Presentation of Results
El Salvador: Data Quality Assessment Framework—Summary Presentation of Results
| Element | National Accounts | Consumer Price Index | Industrial Price Index | Government Finance Statistics | Monetary Statistics | Balance of Payments Statistics | |
|---|---|---|---|---|---|---|---|
| 0.1 | Legal and institutional environment | LO | LNO | LO | LO | O | LO |
| 0.2 | Resources | LNO | LO | LNO | O | LO | LO |
| 0.3 | Quality awareness | LO | LO | LO | LO | LO | LO |
| 1.1 | Professionalism | O | LO | O | O | O | O |
| 1.2 | Transparency | LO | O | LO | LO | LO | LO |
| 1.3 | Ethical standards | O | LO | O | O | O | O |
| 2.1 | Concepts and definitions | LO | O | O | LO | O | O |
| 2.2 | Scope | LNO | O | LO | LO | LO | LO |
| 2.3 | Classification/sectorization | LNO | LNO | LNO | LNO | LO | LNO |
| 2.4 | Basis for recording | LO | O | O | LO | LO | LO |
| 3.1 | Source data | LNO | LNO | LNO | LO | LO | LO |
| 3.2 | Statistical techniques | LNO | LO | LNO | O | O | LO |
| 3.3 | Assessment and validation of source data | LO | O | LO | LO | O | O |
| 3.4 | Assessment and validation of intermediate data and statistical outputs | O | O | O | LO | O | O |
| 3.5 | Revision studies | LO | LNO | NO | NO | LO | LO |
| 4.1 | Relevance | LO | LO | LO | LO | LO | LO |
| 4.2 | Timeliness and periodicity | O | O | O | O | O | O |
| 4.3 | Consistency | LO | O | O | LNO | LO | LO |
| 4.4 | Revision policy and practice | LO | LNO | LO | LO | O | LO |
| 5.1 | Data accessibility | O | O | O | LO | LO | O |
| 5.2 | Metadata accessibility | LO | LO | LO | LNO | LO | LO |
| 5.3 | Assistance to users | LO | LO | LO | LO | LO | LO |
El Salvador: Data Quality Assessment Framework—Summary Presentation of Results
| Element | National Accounts | Consumer Price Index | Industrial Price Index | Government Finance Statistics | Monetary Statistics | Balance of Payments Statistics | |
|---|---|---|---|---|---|---|---|
| 0.1 | Legal and institutional environment | LO | LNO | LO | LO | O | LO |
| 0.2 | Resources | LNO | LO | LNO | O | LO | LO |
| 0.3 | Quality awareness | LO | LO | LO | LO | LO | LO |
| 1.1 | Professionalism | O | LO | O | O | O | O |
| 1.2 | Transparency | LO | O | LO | LO | LO | LO |
| 1.3 | Ethical standards | O | LO | O | O | O | O |
| 2.1 | Concepts and definitions | LO | O | O | LO | O | O |
| 2.2 | Scope | LNO | O | LO | LO | LO | LO |
| 2.3 | Classification/sectorization | LNO | LNO | LNO | LNO | LO | LNO |
| 2.4 | Basis for recording | LO | O | O | LO | LO | LO |
| 3.1 | Source data | LNO | LNO | LNO | LO | LO | LO |
| 3.2 | Statistical techniques | LNO | LO | LNO | O | O | LO |
| 3.3 | Assessment and validation of source data | LO | O | LO | LO | O | O |
| 3.4 | Assessment and validation of intermediate data and statistical outputs | O | O | O | LO | O | O |
| 3.5 | Revision studies | LO | LNO | NO | NO | LO | LO |
| 4.1 | Relevance | LO | LO | LO | LO | LO | LO |
| 4.2 | Timeliness and periodicity | O | O | O | O | O | O |
| 4.3 | Consistency | LO | O | O | LNO | LO | LO |
| 4.4 | Revision policy and practice | LO | LNO | LO | LO | O | LO |
| 5.1 | Data accessibility | O | O | O | LO | LO | O |
| 5.2 | Metadata accessibility | LO | LO | LO | LNO | LO | LO |
| 5.3 | Assistance to users | LO | LO | LO | LO | LO | LO |
IV. Staff’s Recommendations
45. Based on the results of the data quality assessment, discussions with the Salvadoran authorities in the statistics-compiling agencies, and responses from data users, the following actions are proposed to improve El Salvador’s adherence to international statistical standards, to bring the compilation and dissemination of macroeconomic statistics more into line with best practices, and to enhance the analytical usefulness of the statistics.
General recommendations
High priority recommendations
Bring to fruition the new Law on the National Statistics System, establish the new National Statistics Institute, and provide a clear delineation of responsibility for each area of macroeconomic statistics, legal power to collect statistics from all reporters, and sanctions for noncompliance.
Provide a sufficient budgetary, human resource, and information technology resource allocation to agencies to assure the production of essential source data and significant quality improvements in all areas of macroeconomic statistics.
Other key recommendations
Reactivate the Coordinating Committee for the National Statistical System.
Publish the Law of the Central Reserve Bank of El Salvador and the Labor Regulations on the CRBES website.15
Formalize interagency arrangements for collaboration and data sharing.
Introduce ongoing institutionalized quality improvement programs in all agencies, including the use of regular surveys of external users and public advisory groups drawn from all user groups.
Give advance public notice of major changes in methodology, statistical techniques, and data sources.
Improve significantly the public availability of metadata targeted to meet the needs of general and expert users.
Provide the name and contact information of a contact person in each area of statistics to assist users with queries.
National accounts
High priority recommendations
Update the business register and establish a regular program of five-year economic censuses, annual economic surveys, and other statistical products.
Prepare an implementation plan to move to the 1993 SNA and update the reference year.
Other key recommendations
Compile and disseminate more detailed national accounts data, including classifications for government and household final consumption expenditure.
Establish a program to produce basic data and indicators for the activities for which quarterly value added is estimated without indicators.
Value national accounts series at both current and constant prices in U.S. dollars.
Provide more information on data revisions at each stage of the revision cycle.
Provide more detailed documentation on estimation methods for annual national accounts.
Consumer price index
High priority recommendations
Investigate the possibility of using expenditure data from the Annual Multipurpose Household Survey to update the weights for the market basket.
Update the list of market basket items to account for new goods and services.
Update the classification system to the latest five-digit version of COICOP.
Other key recommendations
Eliminate the supplementary monthly CPI produced by the CRBES.
Industrial price index
High priority recommendations
Work with the DIGESTYC to construct a new provisional enterprises sampling frame.
Select an updated enterprise sample.
Update the classification system to revision 3 of the ISIC.
Other key recommendations
Change the name of the IPRI to the producer price index and expand the coverage of the index to include at least mining, manufacturing, electricity, fuel, and water production.
Develop new indices by economic activity as well as by product.
Balance of payments statistics
High priority recommendations
Improve the scope of data sources for compiling holdings of external assets by residents.
Follow the recommendation in BPM5 to sectorize the transactions of the general government.
Compile data on public and private external debt on an accrual basis.
Compile external transactions of the banking sector on the basis of residency criteria.
Make the Data Template on International Reserves and Foreign Currency Liquidity consistent with the classification prescribed in the Operational Guidelines.
Other key recommendations
Improve coverage of communications, construction, and transportation services.
Estimate nonrecorded trade.
Within the current data sources, use market prices in the compilation of the IIP.
Government finance statistics
High priority recommendations
Establish a plan and timetable for adopting the GFSM 2001, commence relevant training, and start recording government transactions on both a cash and an accrual basis.
Classify public institutions using the recommendations of the GFSM 1986 and GFSM 2001 and reconcile this institutional classification with balance of payments, national accounts, and monetary statistics.
Reconcile the operations above and below the line taking into account the differences in the bases of recording for public sector operations.
Other key recommendations
Improve coordination between the units involved in the compilation of government finance statistics within the MF.
Provide for automatic derivation of government finance statistics in the IFAS.
Improve government finance statistics classifications and prepare a functional classification of expenditures and financing data by type of debt holder and financial instrument.
Improve the format of presentation of government finance statistics in the hardcopy publication and on the website.
Monetary statistics
High priority recommendations
Increase staff fully devoted to compiling monetary statistics to accelerate the adoption of the MFSM and guarantee the provision of high-quality data.
Disseminate the weekly central bank and depository corporations surveys.
Complete the adoption of the MFSM, especially with respect to the distinction between resident and nonresident accounts for all financial instruments, sectorization of deposits, loans, and other accounts receivable/payable, valuation of financial assets and liabilities, and treatment of accrued interest.
Ensure consistency in the definition of the general government sector with government finance statistics.
Other key recommendations
Include money market funds in the depository corporations survey.
Since the visit of the mission, the CRBES has posted its law on its own website.
The SDDS is a set of standards, based on best practices, established in March 1996, that guides IMF member countries in the dissemination of their economic and financial data. A detailed description of the SDDS can be found on the IMF’s DSBB at http://dsbb.imf.org.
For further details, see Section III, prerequisites of quality.
That is, to verify not only that the data are released according to the calendar but also that the data disseminated correspond to the metadata posted on the DSBB.
Dissemination practices have continued to be in observance of the SDDS requirements since.
Information on data quality can be found at the IMF website on the Data Quality Reference Site (http://dsbb.imf.org/dqrsindex.htm).
The IPRI is in fact a producer price index that covers only the manufacturing sector.
The legal environment is framed by the Political Constitution of 1983 and comprises mainly the following laws: the Law on the Central Reserve Bank of El Salvador of 1991 and subsequent amendments; the Law on the Superintendency of the Financial System of El Salvador of 1990 and subsequent amendments; the Banking Law of 1999 and subsequent amendments; the Law on Public Financial Administration of 1996 and subsequent amendments; and the Law on the National Statistics Service of 1955. In addition, the following laws and regulations are relevant: the Regulations issued under the Law on Public Financial Administration of 1996 and subsequent amendments; the Labor Regulations of the Central Reserve Bank of El Salvador of 2002; and the Civil Service Law of 1962.
A new Law on the National Statistical System and the Creation of the National Statistics Institute of El Salvador, modeled on international best practice, is currently awaiting presentation to the Legislative Assembly for enactment. A new National Statistics Institute will have technical and administrative autonomy, powers to regulate the NSS, and access to a national budget allocation and other resources. The NSS will be based on centralized setting of standards but decentralized production and dissemination of statistics. However, the Law does not provide a guarantee of sufficient budgetary funding except for censuses.
Passage of the new Law on the National Statistical System and the Creation of the National Statistics Institute of El Salvador would resolve this issue.
The objectives of the programming exercise are specified as the management of balance of payments, fiscal, and financial flows in accordance with the evolution of prices and real production within a framework of macroeconomic stability.
After the visit of the mission, the CRBES began disseminating its Law on its own website.
The introduction of the 1968 SNA and BPM5 methodologies, exceptionally, were announced in advance.
Aware of these shortcomings, the CRBES is making efforts to address some of them in the next publication of the quarterly report.
The Law was published on the CRBES website after the visit of the mission.