## Executive Summary — IMF ROSC data module assessment for France (_cr03339) ## Source details **Canonical URL:** [Executive Summary — IMF ROSC data module assessment for France (_cr03339)](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2003/_cr03339.pdf) ## Other formats - [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2003/_cr03339.pdf.md) - [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2003/_cr03339.pdf.json) --- ### Scope and mission - Reviews France’s data dissemination practices against the IMF’s Special Data Dissemination Standard (SDDS) and provides in-depth assessment of six macroeconomic datasets: national accounts, consumer price and producer price indices, government finance statistics, monetary statistics, and balance of payments statistics. - Producers: - INSEE: national accounts, production indices, labor market, price indices, general government operations (through the general government account), merchandise trade, and population. - Banque de France (BdF): analytical accounts of depository corporations and of the BdF, balance of payments and international investment position, interest rates, and the template on international reserves and foreign currency liquidity. - IMF Statistics Department mission visit: March 5–19, 2003. ### SDDS observance and dissemination practices - France subscribed to the SDDS on August 8, 1996 and began posting metadata on the DSBB on September 27, 1996. - France is in observance of the SDDS since April 27, 2001, meeting specifications for coverage, periodicity, timeliness, and dissemination of advance release calendars. - Data Template on International Reserves and Foreign Currency Liquidity hyperlinked to DSBB on January 24, 2000; National Summary Data Page (NSDP) hyperlinked on November 16, 2001. - Three official institutions compile/disseminate SDDS-prescribed categories: INSEE, BdF, and MINEFI; Euronext compiles share price indices. - Internal government access to data prior to public release disclosed on DSBB; internal access provided not earlier than 24 hours before release to specified ministers for defined categories. - France provided 17 of the 21 summary methodology statements required for SDDS subscription. ### Overall data-quality conclusion - France is in observance of SDDS. - In-depth quality assessment of the six datasets shows strong performance across multiple dimensions: prerequisites of quality, integrity, methodological soundness, accuracy and reliability, serviceability, and accessibility. ### Prerequisites of quality — key findings - Mandates of INSEE and BdF for the six datasets are clearly defined; CNIS addresses reporting burden and confidentiality. - Resources generally adequate; national accounts annual computer system needs updating. - Suggested clarifications: - Clarify INSEE responsibility for government finance statistics relative to MINEFI’s close involvement. - BdF should continue to pursue data sharing with the French statistical system, notably for balance of payments vs. national accounts. - EU/EMU membership has influenced statistical management and increased demand for timely/frequent data. ### Integrity — key findings - Professionalism central at INSEE and BdF; codes of ethics and staff regulations exist (INSEE: civil service code of ethics; BdF: staff regulations and a code of ethics). - Both institutions are transparent in policies/practices and disseminate annual work programs. - Choices of techniques and sources based solely on statistical criteria. ### Methodological soundness — key findings - Six datasets generally follow internationally and/or European accepted methodologies (ESA 95, BPM5, MFSM, Council Regulation (EC) No. 1165/98, etc.). - Noted specific review needs: - Classification and valuation methods in depository corporations statistics. - Valuation method in balance of payments (adjustments for accrued coupons overestimate gross income in selected cases). ### Accuracy and reliability — key findings - Source-data breadth is exceptionally broad; unique identification of economic units facilitates validation. - Statistical techniques and validation practices are sound. - Areas for enhancement: - National accounts: changes in stocks often derived residually; recommend more source data. - PPI: no computerized method systematically applied to detect measurement errors. - Quarterly government finance statistics: finalize by using estimates when accounting data unavailable and conduct revision studies to enhance reliability. ### Serviceability — key findings and recommendations - Relevance, consistency, and timeliness generally satisfactory. - Consider disseminating government finance statistics as a separate statistical product (currently produced as part of national accounts) given strong user interest. - Pursue BdF–INSEE joint task force work to enhance consistency between balance of payments current account and national accounts goods and services account. - Increase transparency of national accounts’ revision policy by aligning quarterly and annual revision periods. ### Accessibility — key findings - INSEE and BdF largely treat statistical products as a public good; many series published free of charge on INSEE and BdF Internet sites. - Methodological notes released for most data; less comprehensive public documentation for government finance statistics (INSEE holds internal documentation that could be adapted for publication). - Publications could better identify the specific INSEE data-producing units as contact points. ### Key timetable and operational facts (verbatim) - IMF mission visit: March 5–19, 2003. - France subscribed to the SDDS on August 8, 1996. - Began posting metadata on the DSBB on September 27, 1996. - In observance of the SDDS since April 27, 2001. - Data Template hyperlinked to DSBB on January 24, 2000. - NSDP hyperlinked to DSBB on November 16, 2001. - The IMF staff began monitoring SDDS subscriber performance in July 2000. - The mission was headed by Ms. Lucie Laliberté and included Messrs. Vincent Marie, Pierre Papadacci, René Piché (all STA), Mrs. Gylliane Gervais and Mr. John Astin (Experts), and Mr. Henry Wright (Administrative Assistant, STA). ### Selected operational recommendations (staff-synthesized) - Clarify INSEE’s responsibility for producing government finance statistics relative to MINEFI. - BdF to continue improving data sharing with INSEE (notably for balance of payments vs. national accounts consistency). - Update the national accounts computer system to meet recent statistical study requirements. - Finalize and improve quarterly government finance statistics by: - Using estimates where accounting data are not readily available. - Conducting and using revision studies to improve provisional estimate reliability. - Implement computerized error-detection methods in PPI validation. - Enhance transparency of national accounts revision policy by harmonizing revision periods for quarterly and annual accounts. - Consider disseminating government finance statistics as a separate statistical product. --- ### Conclusions from the DQAF application to the French statistical system ### Summary assessment - France’s macroeconomic statistics are of high quality: comprehensive, timely, and accurate; adequate for effective surveillance of economic policy. - Detailed assessments use a four-part qualitative scale (details in volume III of the source). ### Legal and institutional environment — highlights - INSEE designated under Articles 32 and 33 of the Budget Law of April 27, 1946; Law No. 51-711 of June 7, 1951 (1951 Statistics Law) central to coordination and secrecy. - France’s decentralized system: 19 government ministries produce statistics as ministerial statistics units (SSMs); CNIS coordinates the annual statistical program. - INSEE role: coordinate methods, centralize documentation, and unify classifications; 1,300 of INSEE’s 7,800 staff are seconded to ministerial statistics units. ### Data sharing, confidentiality, and access - Strong use of administrative sources and cross-referencing (SIRENE, SIE); legal safeguards for confidentiality (1951 Statistics Law, Law 78-17 of January 6, 1978). - Internal governmental access: national accounts submitted to the Office of the Minister at 6:00 p.m. the day before release; public informed via DSBB but not always in French dissemination channels. - CNIS procedures and committees (Statistical Survey Validation Committee, Legal Committee, Statistical Secrecy Committee) provide governance over surveys and confidentiality. ### Resources and systems - National accounts: DCN staff totals and distribution — Seventy in DCN plus another 100 at INSEE and elsewhere; specialized units (e.g., DSE: 45). - Annual accounts computer system is about fifteen years old and inadequate for time-series manipulation, rebasing, and revision studies; urgent need for a new system to support the 2000 base rebasing planned for 2005. - Financial resources adequate for current tasks but may be insufficient for planned reforms (rebasing every five years, expansion, and shortening lags). ### Quality processes and professionalism - INSEE Inspectorate General audits; Quality Charter for Industry Surveys; CNIS questionnaire approval for surveys. - Professional recruitment and training (École nationale de la statistique et de l’administration économique) support high professionalism and ethical standards. --- ### Dataset-specific findings and operational recommendations (selected) ### National accounts - Framework: ESA 95; comprehensive annual and quarterly accounts; annual IO table for 118 industries. - Strengths: - Extensive use of tax and survey data; Système intermédiaire d’entreprises includes individual statements for 100,000 largest nonfinancial firms. - Use of three approaches to GDP (production, expenditure, income) allows cross-validation. - Weaknesses/areas for improvement: - Changes in stocks often estimated residually; recommend greater use of survey/accounting data (quarterly survey of products, costs, and assets with 7,500 enterprises could be used). - Annual accounts computer system needs replacement; resources requested for rebasing to 2000 (planned publication 2005). - Inconsistencies between national accounts rest-of-the-world and balance of payments current account; BdF–INSEE task force recommended. - Priorities: - Allocate resources for a new annual accounts computer system. - Use survey data to measure changes in stocks instead of residual calculations. - Align revision period of quarterly accounts with annual accounts and publicize revision policy. ### Consumer Price Index (CPI) - Coverage: expenditures of all resident households; COICOP classification; CPI basket size: 305 sets of goods and services; approximately 200,000 observations each month. - Practices: - Monthly price survey in all regions during whole month except Corsica. - Definitive CPI is never revised. - CPI detailed data on INSEE site for past 13 months; longer series in aggregate only; INSEE inaugurated indice.insee.fr with access to CPI series back three years (to be completed April 2004); overall CPI series begins in 1970. - Recommendations: - Update methodological guide and disseminate on Internet and in hard copy. - Disseminate longer time series on INSEE website. - Explain utility-price methodology with notional examples. ### Producer Price Index (PPI) - Coverage: approximately 24,000 prices surveyed monthly; product list and weights updated on a five-year cycle; PPI preliminary for six months (generally stabilized after three months). - Weakness: - No computerized method systematically applied to detect measurement errors. - Recommendations: - Establish a computerized method to evaluate potential measurement errors in PPI procedures. - Consider regarding the index as final three months after reference month instead of six months. - New Internet data collection system scheduled for 2004 to include automatic validation modules. ### Government finance statistics - Annual general government accounts conform with ESA 95; classification follows national accounts guidelines rather than GFSM 2001. - Findings: - Annual statistics based on exhaustive accounting sources; quarterly statistics less complete but EU regulation increases demand for quarterly data. - Central government monthly figures on budget operations published 40 days after reference month (10 days beyond SDDS recommendation). - Recommendations/priorities: - Produce a fiscally oriented publication linking national accounts and administrative Treasury data with: - flows and stocks per a government finance framework similar to GFSM 2001; - more detailed revenue and expenditure tables; - cash-based revenue/expenditure/financial flow tables; - bridge tables between government accounting and ESA 95 classifications. - Complete and disseminate general government statistics on a quarterly basis. ### Monetary statistics (BdF) - Framework: broadly in line with MFSM; institutional coverage includes MFIs and UCITS; reporting governed by ECB regulations (e.g., ECB/2001/13). - Strengths: - Comprehensive electronic data collection (BAFI), automated validation, and strong liaison with reporting institutions. - Timeliness: BdF releases central bank data within two weeks after reference date; data for other MFIs within one month. - Issues and recommendations: - Some valuation practices deviate from MFSM (certain securities held to maturity not at market prices); priority to extend market valuation principle and prepare a summary table of deviations from MFSM methodology. - Record all financial derivatives in the balance sheet. - Refine sectorization of claims/liabilities to distinguish central government, central bank, other ESCB components, and other MFIs. ### Balance of payments (BdF) - Framework: BPM5 conformity with some operational deviations (treatment of major construction projects as services; certain "reverse investment" not recorded under direct investment except specific manual cases). - Source data: ITRS complemented by surveys; threshold of 12,500 euros in ITRS introduced following EU recommendation reduced information on households/small businesses; redesign of data collection system spanning 2004 to 2006 expected to remedy deficiency. - Issues: - Adjustments for accrued coupons overestimate gross income from holdings of debt securities for selected transactions. - Discrepancies between balance of payments current account and national accounts rest-of-the-world account remain; BdF–INSEE joint task force in place. - Priorities: - Identify BdF as producer in MINEFI-published tables (implemented end-March 2003). - Revise adjustments for accrued coupons to better measure investment income (implementation: investment income recorded on a full accrual basis from the first quarter 2004 b.o.p release; integrated in information system from 2005). - Pursue integration of balance of payments production into French statistical system and publish revision studies regularly. --- ### Source data collection and processing (BdF) - Monetary statistics: - BdF collects data extensively from depository corporations and other MFIs; MFI sector subdivided into “central bank” and “other depository corporations.” - Reliability enhanced by computerized reporting systems, automated validation modules, quality documentation, and periodic assessment and validation; staff carry out revision studies regularly. - Reporting protocols: mandatory electronic submission, formal/internal/statistical checks on receipt, and systematic follow-up for anomalies; monthly declarations required not later than ten business days after closing date (Decision No. 02-01 of May 22, 2002). - SASM staffing (as of March 15, 2003): 34 regular staff and seven external technical assistants. - Balance of payments source data: - Source: comprehensive ITRS complemented with surveys (DDG, DDP, E84 trade-credit survey), and administrative data (DGDDI customs). - Monthly transaction processing: between 600,000 and 800,000 international transactions each month (reduced after €12,500 threshold introduction on 1/1/2002). - Procedures effectively address usual ITRS problems; redesign of data collection system planned 2004–2006 to recover lost information. - Extensive validation: SAFIR application, formal consistency checks, automatic consistency checks (CCAs), and deeper CCDs; cross-checks with other datasets (BAFI, FIBEN, customs, national accounts). --- ### Transparency and dissemination actions, portal, and timelines - INSEE intends to launch a web portal for official statistics; a first version scheduled to be accessible by the end of 2004. - INSEE commitments and timelines (verbatim excerpts): - New methodological information will be available on the Internet by the end of 2003. A decisive progress will be achieved and noticeable in 2004. - INSEE will implement INSEE Contact (centralized user service); efficiency assessed by the end of 2004. - CPI: CPI series back data to be completed in April 2004; methodological manual to be available on INSEE’s website in 2004. - PPI: internet data collection system scheduled to be implemented in 2004 including automatic data validation and anomaly detection modules. - Balance of payments: investment income on a full accrual basis from first quarter 2004 b.o.p release; integration into information system from 2005 onwards. - BdF identification as producer of balance of payments statistics in MINEFI tables implemented as of end-March 2003. --- ### Staff recommendations (selected, extracted) - General: - Offer a statistical map and actively pursue a statistics portal. - Publish revision studies to promote understanding of statistical revisions. - Inform the public of the precise time government gains access to statistics prior to release. - Increase coordination among agencies dealing with SDDS. - INSEE: - Clearly identify status of statistics (preliminary/semi-final/final) in publications and identify data-producing unit with contact details. - Post more methodological documentation on the Internet. - National Accounts (priorities): - Allocate resources for new annual accounts computer system. - Use survey data on stocks to measure changes in stocks instead of residuals. - Align quarterly-account revision period with annual accounts and publicize revision policy. - CPI/PPI: - CPI: update methodological guide; disseminate longer time series; explain utility-price calculations with notional examples. - PPI: implement computerized error-detection; shorten provisional period from six to three months; develop anomalous shift detection module. - Introduce closer cooperation between CPI and PPI divisions. - Government finance: - Produce a government finance–focused publication with bridge tables and detailed revenue/expenditure tables. - Complete and disseminate quarterly general government statistics. - Monetary statistics: - Extend market valuation principle and record all financial derivatives in balance sheet; refine sectorization of claims/liabilities to distinguish central government, central bank, other ESCB components, and other MFIs. - Balance of payments: - Revise accrued-coupon adjustments; pursue BdF–INSEE task force work; publish revision studies regularly. *Source: Executive Summary and detailed ROSC data-module assessments for France (IMF Statistics Department mission, March 5–19, 2003, and related sections of _cr03339).* ### Executive Summary ...................................................................................................... ### Executive Summary ### Scope and mission - This ROSC data module reviews France’s data dissemination practices against the IMF’s Special Data Dissemination Standard (SDDS) and presents an in-depth assessment of the quality of six macroeconomic datasets: national accounts, consumer price and producer price indices, government finance statistics, monetary statistics, and balance of payments statistics. - Producers of the datasets: - INSEE produces: national accounts, production indices, labor market, price indices, general government operations (through the general government account), merchandise trade, and population. - Banque de France (BdF) produces: analytical accounts of depository corporations and of the BdF, balance of payments and international investment position, interest rates, and the template on international reserves and foreign currency liquidity. - The assessment was carried out by an IMF Statistics Department mission that visited Paris during March 5–19, 2003. ### SDDS observance and dissemination practices - France subscribed to the SDDS on August 8, 1996 and began posting metadata on the DSBB on September 27, 1996. - France is in observance of the SDDS since April 27, 2001, meeting specifications for coverage, periodicity, timeliness, and dissemination of advance release calendars. - The Data Template on International Reserves and Foreign Currency Liquidity was hyperlinked to the DSBB on January 24, 2000; the National Summary Data Page (NSDP) was hyperlinked on November 16, 2001. - Three official institutions compile and disseminate SDDS-prescribed data categories: INSEE, BdF, and MINEFI; Euronext compiles share price indices. - France meets SDDS requirements for the access dimension: advance release calendars are disseminated in hard copy and on INSEE and BdF websites, and a quarter-ahead presentation of release dates is posted on the DSBB. - Internal government access to data prior to public release is disclosed on the DSBB; internal access is provided not earlier than 24 hours before release to specified ministers for defined categories. - France provided 17 of the 21 summary methodology statements required for SDDS subscription; methodological information is disseminated for most SDDS-covered data. ### Main conclusions on data quality (six datasets) - Overall conclusion: France is in observance of SDDS, and the in-depth quality assessment of the six macroeconomic datasets shows strong performance across multiple dimensions. Prerequisites of quality - The mandates of INSEE and the BdF for producing the six datasets are clearly defined; reporting burden and confidentiality receive special consideration via the CNIS. - Suggested clarifications and improvements: - Clarify responsibility of INSEE as producer of government finance statistics because MINEFI is closely involved in preparing those data. - BdF should continue pursuing ways to share data with the rest of the French statistical system, notably for balance of payments with the national accounts. - Resources for producing the six datasets are adequate, although the national accounts computer system needs updating to meet newer statistical study requirements. - Both agencies are fully aware of the importance of quality in data production. - France’s participation in the EU and EMU places it within the European statistical system; legal requirements in the European environment have influenced statistical management and activities (for example, BdF adapting to the transfer of monetary policy-making to the ECB and the creation of the ESCB). Integrity - Professionalism is central to statistical operations at INSEE and BdF. - Both institutions are transparent in policies and practices and disseminate annual work programs. - Codes of ethics and staff regulations exist (INSEE: civil service code of ethics; BdF: staff regulations and a code of ethics). Methodological soundness - The six datasets generally follow internationally and/or European accepted methodologies in concepts and definitions, coverage, classification and sectorization, basis of recording, and valuation. - Specific instances where review would be beneficial: - Classification and valuation methods used in statistics for depository corporations. - Valuation method in balance of payments statistics. Accuracy and reliability - The degree of accuracy and reliability is remarkable in comprehensiveness and level of detail. - Strengths: - Exceptionally broad data sources. - Validation is facilitated by unique identification of economic units across sources and by common concepts and nomenclatures across surveys and administrative records. - Statistical techniques are sound; assessment and validation of intermediate data and outputs are thorough. - Areas for enhancement: - National accounts’ estimation of changes in stocks could use more source data instead of residual derivation. - Producer price index validation could utilize computerized methods to detect measurement errors. - INSEE could finalize the current implementation of quarterly government finance statistics by using estimates when accounting data are not readily available and by relying on revision studies to enhance reliability of estimates over time. Serviceability - French statistics are generally relevant, consistent, and available on a timely basis with required periodicity. - Considerations for improvement: - Given strong user interest in government finance statistics (notably in the EU context), statistical authorities could consider disseminating government finance statistics as a separate statistical product (currently produced as part of national accounts). - Enhance consistency between the current account of the balance of payments and the goods and services account in the national accounts by pursuing the BdF–INSEE joint task force work. - Increase transparency of the national accounts’ revision policy by revising quarterly and annual accounts over the same period. Accessibility - Data sets are accessible to the public; INSEE and BdF largely treat statistical products as a public good. - Methodological notes are released for the data, but less so for government finance statistics (INSEE produces substantial internal documentation that could be adapted for publication). - Data releases could more readily identify the data production units of INSEE as contact points to better meet user demand. ### Key timetable and operational facts (preserved verbatim) - IMF mission visit: March 5–19, 2003. - France subscribed to the SDDS on August 8, 1996. - Began posting metadata on the DSBB on September 27, 1996. - In observance of the SDDS since April 27, 2001. - Data Template hyperlinked to DSBB on January 24, 2000. - NSDP hyperlinked to DSBB on November 16, 2001. - The IMF staff began monitoring SDDS subscriber performance in July 2000. - The mission was headed by Ms. Lucie Laliberté and included Messrs. Vincent Marie, Pierre Papadacci, René Piché (all STA), Mrs. Gylliane Gervais and Mr. John Astin (Experts), and Mr. Henry Wright (Administrative Assistant, STA). ### Selected operational recommendations (staff-synthesized from assessment) - Clarify INSEE’s responsibility for producing government finance statistics relative to MINEFI’s involvement. - BdF should continue to improve data sharing with INSEE and the French statistical system (notably for balance of payments vs. national accounts consistency). - Update the national accounts computer system to meet recent statistical study requirements. - Finalize and improve quarterly government finance statistics by: - Using estimates where accounting data are not readily available. - Conducting and using revision studies to improve estimate reliability. - Improve producer price index validation with computerized error-detection methods. - Enhance transparency of national accounts revision policy by harmonizing revision periods for quarterly and annual accounts. - Consider disseminating government finance statistics as a separate statistical product to meet user demand. *Source: Executive Summary — IMF ROSC data module assessment for France (mission March 5–19, 2003).* ### 15. The conclusions from the application of the DQAF by the IMF to the French ### The conclusions from the application of the DQAF by the IMF to the French statistical system ### Summary - France’s macroeconomic statistics are of high quality. They are comprehensive, timely, and accurate and, thus, are adequate to conduct effective surveillance of economic policy. - Detailed assessments of each dataset are provided in volume III and use a four-part qualitative scale. ### Prerequisites of quality - Legal and institutional environment - INSEE and the BdF conduct statistical operations under the CNIS as well as the European Commission (EC) and the ECB. - CNIS origins: Law No. 51-711 of June 7, 1951, as amended. - On the basis of legal texts that date back to 1946, as well as to the 1951 law, INSEE plays an important role in harmonizing statistical operations. - France’s memberships in the EU and the EMU have confirmed practices (e.g., confidentiality safeguards) and increased demand for timely and frequent data (e.g., quarterly government finance data). - EU/EMU may specify methodological guidelines while leaving substantial room for national practices. - Coordination could be reinforced at the level of statistical outputs given France’s decentralized statistical system. - Institutional roles and responsibilities - INSEE is responsible for national accounts, consumer and producer price indices, and government finance statistics (produced as part of the national accounts). - Responsibility for government finance statistics could be further clarified relative to MINEFI, which closely cooperates in the compilation process. - BdF has the mandate to produce monetary and balance of payments statistics; its authority to collect information from monetary financial institutions stems from the 1984 banking law and the Monetary and Financial Code, further formalized recently within the BdF. - Balance of payments statistics were initially disseminated via a press release of the MINEFI co-signed by the BdF (prior to 2001, it was signed only by the MINEFI). The balance of payments statistical tables on the MINEFI website identified the Ministry as the source, leaving the wrong impression of its involvement in compilation; this situation was fully corrected by end-March 2003. - Resources and systems - Overall, for the four datasets, resources are adequate and procedures indicate quality awareness. - National accounts: trade-off between expertise required and staff mobility for career development could be better balanced. - Insufficient resources are allocated to revamping the computer system for the annual accounts, hindering statistical studies (e.g., consistency of time series and data revisions). Current system: allows consistency checks across variables but does not easily lend itself to comparisons over time, changes in nomenclatures, adjustments, or revision studies. - BdF’s resources are commensurate with the workload of the statistical program. Information systems adequately inform staff of resource needs for statistical action plans; day-to-day monitoring uses efficient computer systems. - Data sharing and reconciliation - BdF should continue to actively pursue ways to share data with the rest of the French statistical system, notably for balance of payments with the national accounts; this could be pursued through the joint BdF-INSEE task force on eliminating discrepancies with national accounts. ### Integrity - Professionalism and ethical standards - INSEE and the BdF maintain a very high degree of professionalism stemming from recruitment and training (notably École nationale de la statistique et de l’administration économique). - Choices of statistical techniques and sources are based exclusively on statistical criteria. - INSEE does not have specific instructions reminding staff to comply with ethical standards, but professionalism and independence are entrenched cultural prerequisites (e.g., Director General of INSEE’s tenure extends beyond the government’s term of office). - BdF: charters of the ESCB and the BdF confirm BdF’s full autonomy; statutory provisions and a recently complemented code of ethics dealing with financial matters guarantee independence of statistical production. No single document summarizes ethical standards, making access more difficult. - Transparency and access - CNIS publishes annual statistical programs and substantial documentation. - Each data table identifies the data producing agency. - Macroeconomic statistics are provided to the government prior to public release and, for aggregate data on monetary financial institutions, to the ECB. - Except via the DSBB, the public is not informed of the government’s access to data before official release, nor of the period of access prior to release. - Changes are generally explained when implemented; sometimes announced in advance (example: conceptual changes of the latest rebasing of the national accounts presented to the press, the CNIS and other users). - Agencies favor a preventive and educational approach with the media rather than rebutting misinterpretation as part of outreach strategy. ### Methodological soundness - National accounts - Conceptual framework: European System of Accounts (ESA 95). - Scope: comprises all prescribed or recommended annual and quarterly accounts, including the complete sequence of accounts for institutional sectors (with financial accounts and balance sheets) at annual intervals; France also builds an annual input-output (IO) table. - GDP estimated through production, expenditure, and income approaches, though a single estimate is published. Annual accounts compilation is essentially based on the IO table; quarterly accounts estimated by a similar method with a smaller IO table. - Delineation of production and assets, classification, sectorization, and basis for recording comply with ESA 95. - French national accounts systematically adjust production and value added upward by activity and sector to take into account tax evasion and informal market activity. - Price indices - CPI: concepts, definitions, and basis for recording follow most ESA 95 principles and EU regulations for the harmonized index of consumer prices. - Scope: expenditures of all resident households. - Classification: COICOP. - CPI basket size: 305 sets of goods and services; annual weights come from the national accounts. - PPI: concepts and definitions align with ESA 95 and Council Regulation (EC) No. 1165/98. - Scope: all resident enterprises engaged in commercial activity; follows EU recommendation to use factory gate prices (excludes changes in inventories). - Excludes own-account production for final consumption, unlawful sales to willing buyers, and services. - Classification of establishments and products by activity follows international standards. - Basis for recording prices and weights: products intended for the domestic market, excluding value added tax and transport costs. - Government finance statistics - Annual statistics of the general government account conform with ESA 95 and the ESA 95 Manual on Deficit and Debt, but are prepared only as part of the government sector in the national accounts—not within a comprehensive and integrated government finance framework. Classifications follow national accounts guidelines rather than those of the Government Finance Statistics Manual 2001. - An integrated government finance framework could be constructed from national accounts components: flows of transactions, other economic flows, and balance sheets as well as debt. - Monetary statistics - Analytical framework broadly in line with the Monetary and Financial Statistics Manual (MFSM). - Classification of financial assets and sectorization generally in line with MFSM, except certain derivatives not recorded in the balance sheet and MFIs not apportioned by type of creditor/debtor (national central bank, other components of the ESCB, or other MFIs). - No detail on claims on governments of BdF and other MFIs, hampering identification of operations with central government. - Basis for recording of flows and stocks broadly follows international standards, except certain securities held through to maturity and issued debt instruments which are not at market prices. - Financial operations recorded on a gross basis inclusive of accrued interest, in accordance with the MFSM. - Balance of payments - Prepared in conformity with BPM5 concepts and definitions. - Major construction-related service projects treated as services rather than direct investment. - “Reverse investment” direct investment transactions usually not recorded under direct investment, except specific cases treated manually. - Highly detailed classification facilitates monthly cross-checks by group of activities/transactions for national accounts purposes. - Authorities value stocks of portfolio investment and direct investment at market prices; direct investment also available at book value. - Adjustments for accrued coupons for bonds overestimate gross income from holdings of debt securities for selected transactions, although net income remains unbiased. ### Accuracy and reliability - National accounts - Data sources (surveys and administrative data) are very comprehensive and exploited systematically. - Système intermédiaire d’entreprises includes individual statements for the 100,000 largest nonfinancial firms and allows for in-depth validation. - Compilation uses sound statistical techniques and sufficiently detailed nomenclatures. - Merchandise trade is deflated using unit value indices because explicit price indices are not generally available. - Changes in stocks are minimally estimated from statistical data and otherwise derived residually. - Quarterly accounts are disseminated in unadjusted and adjusted-for-trading-days-and-seasonal-variations versions, with adjustment techniques applied uniformly. - Use of the three approaches to compile GDP allows staff to further assess and validate data. - Staff use revision studies effectively to improve subsequent estimates. - CPI and PPI source data and processes - CPI source data: exceptionally broad coverage surveys; monthly price survey in all regions during the whole month except Corsica. - Approximately 200,000 observations each month. - Definitive CPI is never revised; technical studies are carried out regularly to improve methodology. - PPI source data: product list and weights updated on a five-year cycle. - Approximately 24,000 prices surveyed monthly, during the whole month. - Products no longer sold are generally replaced with similar products and linked; completely new products incorporated at the next renewal of the list. - No computerized method is applied systematically to detect potential measurement errors in the PPI. - The index is preliminary for six months, although it is generally stabilized after three months. - With source data fully incorporated early, revision studies would not have a statistical impact. - Government accounts - Source data for annual statistics on accounts of general government derived from public accounts for central and local government and from income statements and balance sheets for other central government agencies and social security entities. - Directorate General of Public Accounting of the MINEFI transcribes public accounts data on a national accounts basis; data then adjusted and validated by BdF (financial accounts) and INSEE (nonfinancial accounts). - Agencies adjust accounting data compiled on a cash basis to reflect transactions on an accrual basis. - INSEE assesses and validates data by combining general government accounts with other institutional sectors, ensuring consistency across sectors and statistical areas. - Each year, revisions are accounted for and analyzed for internal use; however, few revision studies are made of long-term series. - If quarterly government accounts are compiled on the basis of estimates for missing accounting data, greater use of revision studies would help ensure reliability of provisional results. *IMF staff conclusions from the application of the DQAF to the French statistical system (conclusions and assessments as presented in the supplied content).* ### 38. The BdF collects source data ### _cr03339 - 38. The BdF collects source data ### Source data collection for monetary statistics - The BdF collects source data for monetary statistics by extensively covering depository corporations and other MFIs. - Data span the entire spectrum of financial assets and economic sectors and are sufficiently detailed to allow an accurate classification of all financial assets and all institutional sectors. - The MFI sector is subdivided into “central bank” and “other depository corporations.” - Statistical techniques are described as sound and yielding reliable results. - Factors contributing to reliability: - computerized systems for source data reporting and processing; - automated validation modules; - the quality of the documentation; - periodic assessment and validation of source data, intermediate data, and final results. - Staff regularly carry out revision studies to identify the main causes of discrepancies between provisional data and final data. ### Balance of payments source data - Source: comprehensive international transactions reporting system (ITRS), complemented with surveys. - Procedures effectively address usual ITRS problems. - Implementation of a threshold of 12,500 euros in the ITRS (following an EU recommendation) has led to a loss of information on transactions of households and small businesses. - A redesign of the data collection system, spanning 2004 to 2006, is expected to remedy this deficiency. - Procedures for assessing and validating source data and intermediate data are exhaustive and systematically applied. - Validation of statistical outputs entails frequent referrals to the source data. - Statistical techniques are well founded; regular revision studies guide choice of collection and adjustment methods. ### Serviceability (relevance, timeliness, periodicity, consistency, revisions) - National accounts (INSEE) - Relevance ensured through CNIS discussions and frequent contacts with government, academia, and business. - No organized forum solely devoted to consulting users of national accounts. - Timeliness: preliminary quarterly accounts released in 50 days, in conformity with the SDDS. - Revision policy: periods of revision of annual and quarterly accounts differ; revision policy not clearly disseminated. - Recommendation: align period of revision of quarterly accounts with that of annual accounts to increase transparency. - Time series based on ESA 95 available at annual and quarterly frequencies beginning from 1978. - Inconsistencies between the rest of the world account in national accounts and the balance of payments. - Publications identify statistics as preliminary, but revised data are not identified as such. - CPI and PPI - Indices meet users’ needs and SDDS requirements for timeliness and periodicity. - No organized user group beyond CNIS discussions; users provide comments via INSEE Information Service. - PPI Internet site receives approximately 20,000 visitors per month. - CPI publication practice: - provisional version limited to few items provided 12 days following the month; - definitive version in full detail is published on the 25th day of the following month. - PPI is revised up to the sixth month following the reference month. - Government finance statistics - Widely used for monitoring deficit and debt criteria for EU purposes. - Suggestions to adapt statistics in a fiscally oriented publication linking national accounts statistics and administrative Treasury data. - Need to improve detail (functional classification, quarterly periodicity). - Statistics are annual and published in full scope at the latest six months after the reference period (in accordance with SDDS). - France’s monthly figures on central government budget operations are published 40 days after the end of the reference month (10 days beyond SDDS recommendation). - Statistics are identified as preliminary, but not as revised. - Monetary statistics (BdF) - Relevance checked periodically for users such as the ECB and BdF’s General Directorate of Economics and International Relations. - Permanent consultation procedure with the ECB and regular procedures with other institutions, banking and financial community, and academia. - Timeliness and periodicity consistent with SDDS: - BdF releases its data within a period of two weeks after the reference date; - disseminates data on other MFIs within one month after the reference date. - Data are internally consistent and consistent across time periods. - Revision policy follows ECB principles and guidelines. - Aggregate position for MFIs (apart from the central bank) is provisional for one or two months. - Users are notified that first data published are preliminary and subject to revision; major revisions and causes are highlighted. - Balance of payments - Annual balance of payments statistical program presented to CNIS committee. - Participation in Commission des comptes et budgets économiques de la Nation and task forces/committees of Eurostat, the ECB, and the EU Commission. - Data timeliness and periodicity meet ECB requirements (more stringent than SDDS). - Balance of payments data are consistent with monetary statistics, but discrepancies remain between the current account and the account for the rest of the world in national accounts; a task force created to harmonize these series. - Annual series broadly in line with BPM5 available beginning from 1946. - BdF publicized the revision policy adopted in 2002 and identifies the status of each series in statistical tables; few revision studies or analyses are made public. ### Accessibility (availability of data and metadata, user assistance) - National accounts (INSEE) - Hard copy publications include commentary, tables, and graphics; CD-ROMs include long series in HTML and Excel. - Some series presented as indices or trends rather than levels. - Release calendars announce releases two months and two weeks beforehand; exact date provided one week in advance. - Data made available to all users simultaneously. - Publications contain metadata, glossaries, and nomenclatures. - Many series, including detailed data, published free of charge on INSEE’s Internet site. - Recent publications do not identify producing unit and contact persons; INSEE centralizes assistance and forwards requests to producing units when applicable. - Annual update of catalog of publications and services, indicating charges and ordering information. - CPI and PPI access - INSEE Internet site: CPI detailed data for past 13 months; longer series in aggregate only; access to longer-term data or metadata not very user-friendly. - PPI: users may easily access both data and metadata for past 36 months on INSEE’s Internet site and in Les Indices de prix de vente industriels et des services aux entreprises. - CPI and PPI disseminated according to a regular calendar; monthly publications contain sufficient detail regarding indices. - User support available; CNIS and PPI website “Indices Pro” indicate names of contact persons, but statistical publications do not. - Government finance statistics - INSEE disseminates general government statistics according to an advance release calendar. - Rapport sur les comptes de la Nation includes methodological explanations; exhaustive sources and methods not systematically released though often obtainable on request. - Contact persons not identified in publications; MINEFI units involved are identified alongside INSEE. - Catalogs provide lists of available documents. - Statistics on INSEE’s Internet site extend from 1978 to the penultimate year available (data for the latest year available only for a fee). - Monetary statistics (BdF) - Readily accessible on BdF Internet site and regular publications. - Calendar of advance release dates published; data disseminated simultaneously to all interested parties. - Nonpublished data made available free of charge. - Metadata accessible via BdF and ECB statistical publications and Internet sites. - User support is rapid and efficient; each publication and table includes contact information for the producing unit. - Catalog of publications available on BdF Internet site and in complimentary hard-copy form. - Balance of payments (BdF) - BdF Internet site contains recent edition of Bulletin de la Banque de France, annual report La balance des paiements et la position extérieure de la France, and detailed methodological notes. - Free access to numerous balance of payments series and an advance release calendar. - Data and metadata also available on the MINEFI website. - Annual report provides quality presentation and analysis facilitating understanding of developments. - User support prioritised: each statistical table identifies the unit responsible and a telephone number; Internet site logs requests for information. - Catalog of BdF publications published annually with product costs. ### Staff recommendations (to strengthen conformity and analytical usefulness) - General - Offer a statistical map to help users identify available statistical outputs; actively pursue a statistics portal. - Promote broader understanding of statistical revisions by publishing revision studies. - Inform the public that the government has access to statistics prior to their publication, specifying the precise time of access. - Increase coordination for all agencies dealing with the SDDS. - INSEE - Ensure status of statistics (preliminary, semi-final, or final) is clearly identified in publications. - Ensure data-producing unit is identified in publications, including a telephone number and e-mail address. - Post more documentation on methodologies for macroeconomic series on the Internet. - National Accounts - Priorities: - Allocate greater resources for implementing a new computer system for annual accounts. - Make greater use of survey data on stocks to measure changes in stocks, instead of calculating them residually, in the annual and quarterly accounts. - Align period of revision of the quarterly accounts on that of the annual accounts and make the revision policy known to the public. - Other recommendations: - Expedite production of explicit price indices for exports and imports of goods. - Improve breakdown of intermediate consumption by product. - Explore making quarterly accounts more consistent with annual accounts with respect to volume measurement. - Review deflation of exports and imports of services, making better use of balance of payments source data. - Adopt a more structured approach toward consultations with users. - Consumer Price Index - Priorities: - Update the methodological guide on the consumer price index in hard-copy form and on the INSEE website. - Disseminate longer time series on the INSEE website. - Disseminate information on anticipated major changes in advance. - Envisage publicly identifying more details in the commodities selected (variety). - Explain in public methodology how utility prices are calculated, using notional examples. - Producer Price Index - Priority: - Establish a computerized method for evaluating potential measurement errors in PPI procedures, to be used systematically. - Other recommendations: - Regard the index as final three months after the reference date, instead of six months. - Explore developing a module for detecting anomalous price shifts in the new automatic price collection system. - Consumer and Producer Price Indices - Priority: - Introduce closer cooperation between the two price divisions to exploit synergies. - Government Finance Statistics - Priorities: - Produce a publication focusing on government finance including: - tables of flows and stocks according to a “government finance” framework similar to Government Finance Statistics Manual 2001; - more detailed revenue and expenditure tables; - a table on cash-based revenues, expenditures, and financial flows; - bridge tables between government accounting data and the account of the central government (“État”) subsector; - a bridge table between tax and budget nomenclatures, and ESA 95 classifications. - Complete and disseminate statistics on general governments on a quarterly basis. - Identify assistance to users more clearly. - Monetary Statistics - Priorities: - Extend general principle of market valuation for balance sheet items; prepare detailed summary table of deviations from MFSM methodology. - Record all financial derivatives in the balance sheet. - In surveys for the BdF and aggregate survey for other resident MFIs, refine sectorization of claims on and liabilities to all MFIs to distinguish central government, central bank, other components of the ESCB, and other MFIs. - Balance of Payments - Priorities: - Identify the BdF as the producer of the statistics in statistical tables published by the MINEFI.6 - Revise adjustments for accrued coupons (“coupons courus”) on transactions in bonds to better measure investment income. - Pursue efforts of joint DBDP-INSEE task force to achieve greater coherence between balance of payments current account and the rest of the world account in the national accounts. - Other recommendations: - Pursue integration of balance of payments production process into the French statistical system through exchanges of source data. - Publish revision studies and analyses regularly. - Footnote: - 6 Implemented as of end-March 2003. *Source: _cr03339 - 38. The BdF collects source data* ### 1.2 Transparency ### 1.2 Transparency ### Overview of assessment indicators - The assessment table in the source shows the ROSC Data Module coverage across multiple dimensions including Transparency, Ethical standards, Methodological soundness, Accuracy and reliability, Serviceability, and Accessibility. - For Transparency specifically, the row is labeled "1.2 Transparency" with subsequent columns showing status indicators "O" (observed) across six columns in the printed table. ### Context and responsibilities - Reactions concerning balance of payments and monetary statistics have been written by the Banque de France. - Other reactions come from INSEE. ### Key recommendation on dissemination and coordination - Offer a statistical map to help users identify the available statistical outputs; in this regard, actively pursue the idea of a statistics portal. - Rationale: In a decentralized statistical system, coordination is just as important for data dissemination as for data collection. - INSEE intends to facilitate users’ search for statistics on the web. - INSEE statement: "In order to reach this objective, a project of web portal for official statistics will be launched at the beginning of" *Source: Report on the Observance of Standards and Codes (ROSC)—Data Module; Response by the Authorities; October 21, 2003* ### 2004. A first version of this portal is scheduled to be accessible to users by the end of 2004. ### 2004. A first version of this portal is scheduled to be accessible to users by the end of 2004. ### General transparency and dissemination actions - Promote a broader understanding of statistical revisions by publishing revision studies to familiarize users with the workings of the statistical process. - INSEE regularly conducts revision studies, some being published; a further step will be taken to publish revision studies. - The Banque de France will increase communication on major revisions in its publications. - Inform the public that the government has access to statistics prior to their publication, specifying the precise time at which the government gains access thereto. - INSEE already gives a precise description of Government’s access to statistics prior to their publication, including the exact timing and identity of the relevant members of Government, in the Dissemination Standards Bulletin Board of the IMF. - INSEE will also provide this information on its own web site to improve transparency vis-à-vis the general public. - Concerning Balance of Payments statistics, the Banque de France and the Treasury will keep endeavoring to reduce to the minimum needed the time frame between the availability of information and its public disclosure. - Increase coordination for all agencies dealing with the SDDS. - INSEE will inform the IMF of all updates of the release calendar and the SDDS data web page. - A network between all agencies dealing with the SDDS will be activated for instantaneous sharing of information. ### INSEE operations, user contacts, and methodology documentation - Ensure that the status of statistics (whether preliminary, semi-final, or final) is clearly identified in publications. - Publications on price indices already mention the status of the data. - A clarification for national accounts has been made as part of a new communication policy. - Ensure that the data-producing unit is identified in publications including contact details. - INSEE argues that providing telephone numbers and e-mail addresses for all producing units would create numerous basic requests better handled centrally. - INSEE implemented a centralized service “INSEE Contact” which: - collects all requests of information, - answers standard basic ones directly, - transmits non-standard requests to the right contact persons among data-producing units and tracks the whole answering process. - Efficiency of the new service will be assessed by the end of 2004. - Post more methodological documentation on the Internet for macroeconomic series. - Existing internal methodological documentation must be adapted for a broader audience; summaries are being prepared. - New methodological information will be available on the Internet by the end of 2003. A decisive progress will be achieved and noticeable in 2004. ### National accounts — priorities and responses - Allocate greater resources for implementing a new computer system for annual accounts. - New developments achieved; part of the new system is already implemented. - A temporary high level statistician function has been budgeted by INSEE to further reorganize a more consistent and integrated system. - Make greater use of survey data on stocks to measure changes in stocks instead of calculating them residually. - A specific survey for change in stocks has been conducted for many years but results were weak and very volatile. - Regular use of annual business accounting data led to a first progress. - A new test for possible use of the quarterly survey will be completed as soon as it is renovated (second semester of 2003). - Align period of revision of quarterly accounts on that of annual accounts and make revision policy known to the public. - The sum of unadjusted data of the four quarters of a year is maintained equal to the corresponding annual data. - INSEE will not adopt the IMF suggestion that quarterly profiles should not be revised after annual data are definitive, because keeping old quarterly profiles unchanged would damage the quality of recent years. - INSEE recognizes revision policy should be clearer to the public: - Information on revisions and dissemination policy (including delays and contents) was added in the CD-Rom published in June 2003. - Shorter explanations on revisions included in quarterly press releases since the end of September 2003 and will be added in the next annual press release (April 2004). ### National accounts — other recommendations - Expedite production of explicit price indices for exports and imports of goods. - Explicit indices for exports of goods started from September 2001; whole industry should be covered within 3 to 4 years. - Work on indices for imports will start in the beginning of 2004. - Program is very costly and difficult to expedite. - Improve breakdown of intermediate consumption by product. - Progressive improvements using different sources on specific products; a general survey is not appropriate due to statistical burden concerns. - Explore possibility of making quarterly accounts more consistent with annual accounts with respect to volume measurement. - Implementation taken into consideration in INSEE’s work program for 2004. - Review deflation of exports and imports of services using balance of payments source data. - INSEE will not follow this recommendation because appropriate data do not exist in the balance of payments source. - Adopt a more structured approach toward consultations with users. - INSEE is implementing a permanent organized forum devoted to consulting users of national accounts in addition to CNIS discussions. - The forum will be organized through electronic discussions between annual meetings. The first meeting is to be organized by the end of 2003. ### Consumer Price Index (CPI) - Update methodological guide on the consumer price index in hard-copy form and on the INSEE website. - INSEE added a complete review of documentation disseminated, including broader use of Internet, in its mid term work program. - Disseminate longer time series on the INSEE website. - INSEE inaugurated “indice.insee.fr” which offers access to all CPI series back to three years ago; these back data will be completed in April 2004. - The overall consumer price index is available at "insee.fr" website; the series begins in 1970. - Disseminate information on anticipated major changes in advance. - Efforts will be made in the close future. - Envisage publicly identifying more details in the commodities selected (variety). - INSEE notes disclosure rules balance statistical confidentiality and prevention of external intervention; INSEE intends to reassess justification of these rules. - Explain method used to calculate utility prices with notional examples in public methodology. - The already published methodological reference manual will be available on INSEE’s website in 2004. ### Producer Price Index (PPI) - Establish a computerized method for evaluating potential measurement errors in PPI procedures to be used systematically. - Currently, business answers are manually checked by sector statisticians who contact businesses when anomalies arise. - A computerized check can only be considered in a full redesign of the computer software package. - The new data collection system via Internet scheduled to be implemented in 2004 will incorporate an automatic data validation module allowing automatic checks of price changes; the recommendation will be progressively followed from 2004. - Regard index as final three months after the reference date instead of six months. - Production price indices currently subject to change during a six month period due to late answers and changes of weights. - Considering very minor revisions recently, INSEE plans to reduce the period of revision down to three months. - Explore developing a module for detecting anomalous price shifts in the new automatic collection system. - The 2004 Internet data collection system will include an automatic data validation module, including automatic checks of price shifts, for sector statisticians (not for respondents) to avoid bias. ### Consumer and Producer price indices — cooperation - Introduce closer cooperation between the two price divisions to exploit synergies. - Divisions already participate in monthly meetings of the Indices Workshop. - Planned cooperation: - Econometric work to calculate hedonic estimations of computer prices. - Intensified cooperation on prices in business services and agriculture sectors. - Continued joint work on energy sectors. ### Government Finance Statistics - Produce a publication focusing on government finance comprising: - tables of flows and stocks according to a “government finance” framework similar to the one in the Government Finance Statistics Manual 2001; - more detailed revenue and expenditure tables; - a table on cash-based revenues, expenditures, and financial flows; - bridge tables for the main aggregates between government accounting data and the account of the central government (“État”) subsector; and - a bridge table between tax and budget nomenclatures, and ESA 95 classifications. - INSEE will work in this direction and create a specific domain devoted to general government accounts on INSEE’s website in 2004, containing metadata and bridge tables on main annual data on budget revenues and expenditures. - INSEE will probably not assemble all fiscally oriented data suggested due to lack of consistency across data (differences in periodicity and dissemination delays) that could confuse users. - Complete and disseminate statistics on general government on a quarterly basis. - This has been part of INSEE’s plans for two years as enforcement of the European regulation. - Quarterly accounts on general government will be disseminated as soon as quality is ensured; these data are currently being tested for the first time. - Identify assistance to users more clearly. - INSEE will cautiously follow this recommendation; the name and phone number of a contact person will only be indicated on the CD-Rom publication. ### Monetary statistics - Extend principle of market valuation for balance sheet items and prepare a detailed summary table of deviations from MFSM methodology. - DESM has used bookkeeping sources to stay coherent with financial institutions’ accounting standards. - With implementation of International Accounting Standards in Europe, INSEE will extend market valuation where appropriate and regularly update a detailed summary table of deviations from MFSM methodology. - Record all financial derivatives in the balance sheet. - Plans to extend the scope of financial derivatives included in balance sheet statistics in connection with ongoing Eurosystem work. - In surveys for BdF and aggregate surveys for other resident MFIs, refine sectorization of claims and liabilities to distinguish between central government, central bank, other components of the ESCB, and other MFIs. - This distinction is available for quarterly data and will soon be published. ### Balance of Payments - Identify the BdF as the producer of the statistics in statistical tables published by the MINEFI. - Implemented since end March 2003 (indicated in a footnote). - Revise adjustments for accrued coupons (“coupons courus”) on transactions in bonds to better measure investment income. - Implementation based on conclusions of a joint ECB-Eurostat task force. - Investment income will be recorded on a full accrual basis from the first quarter 2004 b.o.p release and integrated in the information system from 2005 onwards. - Pursue efforts of the joint DBDP-INSEE task force to achieve greater coherence between the balance of payments current account and the rest of the world account in the national accounts. - The task force is a top priority for INSEE (DCN) and Banque de France (DBDP) and an European objective for elaborating quarterly national accounts on a sectoral basis. - Initial focus on transportation (where bias is most significant) and bringing together production timetables. - Pursue integration of balance of payments production process into the French statistical system through exchanges of source data. - Such exchanges already exist with the Tourism Directorate and with INSEE (DCN) and will be intensified. - Publish revision studies and analyses regularly. - DBDP publishes several versions of its data with clear identification of status (provisional, semi-definitive or definitive) but has not regularly communicated reasons for revisions except for important methodological changes. - DBDP intends to enhance communication on revisions in the annual report and on a monthly basis when revisions are significant for interpretation. ### Final comment regarding the ROSC - Suggestion from French authorities preparing the ROSC-data module: - Taking into account the importance in Europe of the general government debt (calculated according to the Maastricht treaty definition), the IMF should consider its inclusion in the SDDS set of data on general government operations. *International Monetary Fund — FRANCE: Detailed Assessments Using the Data Quality Assessment Framework (DQAF), Prepared by the Statistics Department; Approved by Carol S. Carson and Michael C. Deppler; October 21, 2003.* ### 1. DQAF: Summary of Results for National Accounts Statistics............................................42 ### 1. DQAF: Summary of Results for National Accounts Statistics............................................42 ### Overview - The assessment was prepared by an IMF Statistics Department mission that visited France from March 5 to 19, 2003. - The detailed information was gathered from publicly available documents and information provided by French officials, organized using the generic DQAF (see Appendix II). - Four datasets assessed are produced by INSEE: national accounts, consumer prices, producer prices, and government statistics. Monetary statistics and balance of payments are produced by the Banque de France (BdF). - To avoid repetition, the assessment groups the DQAF dimensions "Prerequisites of Quality" and "Integrity" for INSEE in the national accounts section and for the BdF in the monetary statistics section, with supplements as required in related datasets. ### Prerequisites of quality — Legal and institutional environment - INSEE is a Directorate General of the Ministry of the Economy, Finance and Industry (MINEFI) and is designated as France’s statistics producing agency under Articles 32 and 33 of the Budget Law of April 27, 1946. - Article 1 of the implementing decree specifies INSEE’s tasks include “collecting, compiling, and updating statistics relating to the state and flows of people and goods...” - France’s statistical output is decentralized: 19 government ministries produce statistics as ministerial statistics units (SSMs). - Decree 84-628 of July 17, 1984 (Article 2) defines producing agencies to include INSEE, SSMs, government agencies, public entities, and private entities providing public services or collecting economic and social data with government financing or consent. - Reference legislation includes: - Law 51-711 of June 7, 1951 (the 1951 Statistics Law), as amended, on statistical reporting requirements, coordination, and secrecy; - Decree 84-628 of July 17, 1984; - Order of December 12, 1997 on the operating procedures of the Statistical Survey Validation Committee; - Law 78-17 of January 6, 1978 on informatics and freedom. - INSEE was given responsibility for the national accounts in 1965; other MINEFI units and the BdF are closely involved, but the INSEE National Accounts Department (DCN) defines methods and is accountable for results. - France participates in the European Statistical System; Article 285 of the Treaty of Amsterdam and Regulation EC 322/97 of February 17, 1997 frame national and European responsibilities for statistics. ### Prerequisites of quality — Data sharing and coordination - Article 1 of the 1946 Decree assigns INSEE the role “... to coordinate the statistical methods, resources and work of government agencies and private entities ... to centralize their statistical and economic documentation and to unify the statistical classifications and codes...” - INSEE supports coordination by seconding staff: 1,300 of its total staff of 7,800 are seconded to ministerial statistics units. - Procedures for sharing and coordinating data are governed principally by the 1951 Statistics Law and Decree 84-628 of July 17, 1984. - The MINEFI compiles an annual program of official statistical surveys, prepared by the CNIS, under which INSEE and SSMs submit their survey plans. - The CNIS reviews the statistical program using three committees: - The Statistical Survey Validation Committee (Article 12-1 of Decree 84-628) determines surveys in the general interest. - The Legal Committee (Article 7 of the 1951 Statistics Law) deals with failures to respond to statutory surveys. - The Statistical Secrecy Committee handles access requests to enterprise databases. - The CNIS secretariat is provided by INSEE; the General Director of INSEE is responsible for carrying out the Decree setting the program of statistical surveys for the current year, published in the Journal Officiel. - The INSEE DCN coordinates compilation of national accounts with external units, issues a detailed timetable for annual accounts in October (following a draft timetable sent in July), and adjusts it as needed. ### Prerequisites of quality — Confidentiality and use of respondent data - The 1951 Statistics Law safeguards statistical confidentiality and use of data solely for statistical purposes: - Article 6: individual statistical information relating to personal and family life may not be communicated until one hundred years after the census or survey. - Article 2: information of an economic and financial nature may not be used for tax audits or monetary sanctions. - Article 7bis of the 1951 Statistics Law: administrative data transferred for statistical purposes may be used by INSEE and SSMs solely for compiling statistics. - Law 78-17 of January 6, 1978 (establishing CNIL) subjects transfers of personal information to strict confidentiality safeguards. - The Statistical Secrecy Committee may grant access to individual survey data when purposes are legitimate (Article 7bis), only to parties that commit not to disclose them (Article 22 of the 1984 Law). - Individual statistical data may circulate within the official statistics system as needed, subject to coordination supervision; disclosure requires a joint decision by the Ministers concerned and CNIL approval when applicable. - Staff access obligations: - Staff with access to confidential data are bound by Article 7bis of the 1951 Statistics Law and by the obligation of professional secrecy established by Law 83-634 of July 13, 1983 (Article 26). - Criminal penalties apply under Article 226-13 of the Criminal Code for breaches of professional secrecy (one year’s imprisonment and a fine of FRF 100,000). - CNIS has general rules for disclosure of aggregated enterprise data and case-by-case criteria for aggregated data on individuals; CNIL’s criteria on legal protection of personal data are more restrictive than CNIS’s. - European-level strengthening: Directive of the European Parliament and of the Council of October 24, 1995 on protection of individuals regarding personal data processing; Council Regulation of June 11, 1990 on transmission of data subject to statistical confidentiality to the Statistical Office of the European Communities (as amended). *Source: DETAILED ASSESSMENTS OF FRANCE’S MACROECONOMIC STATISTICS USING THE DATA QUALITY ASSESSMENT FRAMEWORK (DQAF), IMF mission visit March 5–19, 2003.* ### Annex 21 of the Treaty on the European Economic Area. ### Annex 21 of the Treaty on the European Economic Area ### Legal mandate for statistical reporting and survey approval (0.1.4) - Article 3 of the 1951 Statistics Law makes responding to surveys officially approved (i.e., those included in the statistical program of the CNIS) mandatory. - The CNIS Legal Committee provides its opinion on imposing fines in the cases covered by Article 7 for natural and legal persons who fail to respond after receiving an official warning, or deliberately provide an inaccurate response. - Statutory requirement to respond is paired with concern for limiting respondent burden; some surveys are optional but still examined by CNIS to certify quality and potentially grant status of a “survey in the general interest.” - The CNIS annual program includes ongoing surveys and “new” surveys (new projects, renovated surveys, surveys with cycles longer than one year, or surveys that have not been examined by the CNIS for at least five years). - Inclusion of a survey in the program requires two rulings: - First ruling: examination in one of the thirteen specialized groups of the CNIS committee. If approved, the appropriate group of the Survey Validation Committee delivers a ruling certifying usefulness, that it is in the general interest, non-redundancy, and necessity of the survey. For regional surveys, the Regional Committee for Economic and Social Information, or a duly authorized regional body, must approve. - Second ruling: examination by the Survey Validation Committee to assess quality of planned implementation procedures. If assessment is favorable, the Committee issues a certificate of “compliance” with statistical standards, approves the survey as being in the general interest and of statistical quality, and, if necessary, proposes a ruling to make responding mandatory. - For a survey to be deemed in the general interest, it must meet criteria relating to dissemination, methodology, sample size, questionnaire contents, and nature of results. - Approved surveys are authenticated in the Journal Officiel, providing assurance of usefulness, rigor, and quality; CNIS coordination, committee membership, published deliberations, and a single list of surveys help limit respondent burden and prevent redundancy. ### Data exchange and large-enterprise access - Data exchanged with units outside of INSEE for compiling the national accounts are usually aggregated and transmitted via e-mail in the form of Excel tables. - Data on the 100,000 largest enterprises can be accessed via the database of the Intermediate Enterprise System (SIE, see 3.1.1). ### Staff resources and organization (0.2.1) - Total staff participating in compiling the national accounts: Seventy people working in the DCN, and another 100 people at INSEE and elsewhere. - Specialized units and staff counts: - Public Accounting Directorate General (DGCP) of the MINEFI: specialized unit (19 persons) for transcribing government accounts into the national accounts. - Forecasting Directorate (MINEFI): unit (14) for compiling estimates of preliminary general government statistics and the quarterly accounts pending provision of accounting data by the DGCP. - Monetary and Financial Statistics Directorate of the BdF: unit (12) for compiling financial and nonfinancial banking statistics and the flow-of-funds table for all institutional sectors. - Enterprise Statistics Directorate (DSE) at INSEE: (45) for compiling most nonfinancial corporation and unincorporated enterprise statistics and the supply and use tables for each product, along with trade and transport margins. - Ministerial statistics units: (25) for compiling satellite accounts and providing data on health and housing. - INSEE has no plans to provide additional permanent resources despite the possibility of obtaining extra temporary resources. - INSEE staff classification and training: - Staff divided into four categories, from A to D; more than half of the DCN staff members are in category A. - Continuing education provided to all categories throughout careers; management and supervisory personnel (category A) required to change jobs after their first three years and again after another three years. - Staff turnover is high and creates challenges for program management at the DCN. - Recruitment largely through competitive exams and training in INSEE schools (see 1.1.1). ### Computer resources (0.2.1) - DCN has two computer systems: one for the annual accounts and another for the quarterly accounts. - Quarterly accounts computer system: has a database and applications for manipulating time series and is generally considered to perform very well. - Annual accounts computer system: about fifteen years old; database is made up of tables and not time series, causing: - Difficulty comparing with previous years and complications for the Quarterly Accounts Division needing regular access. - Limitations became apparent when accounts were rebased to 1995. - Delays or prevention of dissemination of some previously available tables. - Need for parallel measures (e.g., transfer of series to more user-friendly systems) to work on time series. - Inability to meet demands such as storage of adjustments to ensure transparency, revision studies, and presentation in new classification formats. - Urgent need for a new system providing storage and manipulation of time series; without more resources, the rebasing of the accounts to 2000, planned for 2005, could be jeopardized. ### Financial resources (0.2.1) - Resources allocated for compiling the national accounts are adequate at present. - Resources may be insufficient to implement planned reforms and expansion, including: - Rebasing that takes place every five years and related statistical work. - Shortening of the lags for compiling quarterly accounts to dissemination 30, 45, and 90 days after the end of the reference period. ### Measures to ensure efficient use of resources (0.2.2) - Annual and medium-term work programs are drawn up for each INSEE unit to ensure coordination and optimal use of resources. - Planning process: - Activities identified at divisional level by activity program, rolled up at Department level, resulting in detailed program for INSEE. - INSEE’s annual work plan submitted to CNIS for discussion and approval; two stages: draft program in February and definitive program in October of the previous year. - DCN subject to periodic audits by the INSEE Inspectorate General to assess efficiency. - Productivity gains in other parts of INSEE (e.g., enterprises filing tax returns on line starting in 2004; use of notepad PCs to collect prices for CPI) have produced indirect benefits for DCN. ### Quality awareness and quality processes (0.3) - INSEE’s Inspectorate General periodically assesses statistical work; main 1999 audit conclusion: preliminary annual statistics and fourth-quarter quarterly statistics should be produced simultaneously and coherently to avoid redundant work; recommendation implemented. - Audit recommendations followed up with compliance assessments in June 2000 and November 2001. - Human resources emphasis on recruitment and training via unified training system (see 1.1.1) and continuing education through courses, seminars, and internal mobility. - Quality Charter for Industry Surveys introduced March 19, 2001, by the “enterprise configuration” of the Survey Validation Committee: - Charter applicable to all surveys conducted by INSEE or an authorized industry association. - Provides for relevancy, accuracy, comparability with other sources, accessibility, and prompt dissemination of results. - Covers questionnaire design, sample selection, data collection and verification procedures, nonresponse estimates, resources provided for the survey, timetables, statistical secrecy, and dissemination of results. - Verification indicators at each step in compiling national accounts: - Tax and survey data on large enterprises are crosschecked. - Contributions of basic indicators to quarterly changes in GDP are measured. - Changes in data sources are assessed for each major dataset in the annual accounts. - Requirement that each questionnaire be approved by the CNIS ensures questionnaires meet statistical program requirements and have been reviewed by the Survey Validation Committee. - CNIS provides forum for debating new and sensitive topics (e.g., poverty and social exclusion, insurance, nonprofit associations). - Inspectorate General promotes implementation of INSEE’s strategic objectives and notifies General Director and Management Board of concerns about official statistics and quality/efficiency. ### Professionalism and integrity (1.1) - 1951 Statistics Law provides legal framework guaranteeing INSEE impartiality. - European Commission decision requires Community statistics to conform to impartiality, reliability, objectivity, and scientific independence (see 0.1.1). - INSEE is a Directorate General of the MINEFI; only appointment by the Council of Ministers is the General Director. Tradition of General Directors’ tenure not coinciding with government term upholds independence. - INSEE has its own web address (insee.fr), which makes no reference to the Ministry overseeing it. - Staff professionalism reinforced by recruitment largely from INSEE’s Group of Higher Institutes of Economics and Statistics, which provides training in statistical ethics and professional conduct. - Choices of sources and statistical techniques are based solely on statistical criteria, including application of international standards (1993 SNA, ESA 95); Eurostat guidelines applied when uncertainty exists (e.g., estimates for indirectly measured financial intermediation services). - INSEE engages in public discussions only rarely; in cases of erroneous interpretation or misuse, INSEE favors bilateral approach with journalists via the “Bureau de presse” of the “INSEE Info-Service” Department and emphasizes preventive and educational outreach rather than invoking right of reply. ### Transparency (1.2) - INSEE website provides access from home page to statutes of INSEE, missions, relevant legal references, and statistical confidentiality clauses; access to European legal documents related to statistics is also provided. - September 2001 edition of Courrier des statistiques provides a complete description of the French public statistics system (update of June 1992 review), explaining role of INSEE and SSM, legal and institutional framework, organization and coordination issues, and system of schools. - CNIS website contains extensive information on France’s statistical system, including reference texts on French and European legal texts, information on surveys, and 5-year statistical programs by statistical domains; program of official surveys is published in the Journal Officiel. - Internal governmental access prior to release: - National accounts are submitted to the Office of the Minister at 6:00 p.m. on the day before their release and released to the press at 8:45 a.m. the next day. - INSEE locks up documents under preparation so only data producing units have access, to prevent leaks. - Public is informed about government’s prior access through metadata disseminated on the IMF’s SDDS website but not in INSEE’s publications or website. - Data releases are clearly identified as INSEE’s product; each data table mentions INSEE as the source. - Advance notice of major changes: public usually notified when changes implemented; major changes sometimes announced in advance (e.g., rebasing coinciding with adoption of 1993 SNA and ESA 95 concepts, with presentations for press, CNIS, and users and publications in Courrier des statistiques and Economie et statistiques). ### Ethical standards (1.3) - All civil servants and government agents bound by obligation of professional secrecy (see 0.1.3) and discretion, requiring refraining from criticizing the institution even after departure. - Laws 25-2 and 25-3 of July 15, 1982 and Law 93-122 of January 29, 1993, as amended, set up ethics commissions governing relations between civil servants and the private sector; laws and regulations provide for penalties for violations of the code of ethics. - Ethical rules relating to professional secrecy and statistical secrecy do not seem systematically presented to staff in a document or individual letters except for new arrivals. - Survey questionnaires all mention statistical secrecy and staff working on individual data are very conscious of this requirement. *Annex 21 of the Treaty on the European Economic Area — (source content provided).* ### 2. Methodological soundness ### 2. Methodological soundness ### 2.1 Concepts and definitions - The 1993 SNA and ESA 95 form the framework used to compile the national accounts. - The definitions of the constituent units of the economy follow the 1993 SNA. - Institutional sectors include nonprofit institutions serving households (NPISHs). - The basic statistical unit for production is the "unit of homogenous production" (UHP), derived from a statistical breakdown of the enterprise; local establishments are not the basic unit because they do not keep full accounts. ### 2.2 Scope - Annual accounts compiled since 1959; quarterly accounts compiled since 1978. - System covers annual and quarterly variables as follows. - Annual statistics: - Value added and GDP at current prices and at the previous year’s prices, for 118 industries; - The expenditure measure of GDP at current prices and the previous year’s prices; - A breakdown of value added by industry in an operating account at current prices; - The complete sequence of accounts for the institutional sectors, including financial accounts and balance sheets, along with the rest-of-the-world accounts; - An annual input-output table (IOT) for 118 industries by 118 products (but 480 products for the supply and use tables). - Quarterly statistics: - Value added and GDP at current and constant prices (1995=100) for 39 industries (but 49 for production), presented in groups of 16 for publication purposes, unadjusted data and data adjusted for working days and seasonal variations (WD-SA); - The expenditure measurement of GDP at current and constant prices (1995=100), unadjusted data and data adjusted for WD and SA; - Production, operating, and income accounts for industrial sectors; - Operating accounts by industry (16 industries). - Coverage: - Production boundaries follow the 1993 SNA and include: - Own-account production of goods for own final consumption; - Own-account research and development activity; - Production of goods for own-account fixed capital formation; - Mineral exploration; - Production of entertainment, literary or artistic originals; - Production of computer software; - Illegal output sold to willing buyers. - In practice, intra-enterprise deliveries for intermediate consumption are rarely recorded, causing underestimation of production and intermediate consumption (but not value added) and affecting the breakdown of value added by industry. - Production, marketing, and consumption of illegal substances (e.g., narcotics) are not estimated. - Systematic adjustments are made for undeclared goods and services (undeclared employment and understatement of earnings), which totaled 5 percent of GDP in 1985 and 3 percent in 2000. - Asset boundaries are consistent with the 1993 SNA, except for works in progress in agriculture and forestry; plans call for replanted forests to be counted as assets when the national accounts are rebased (2000 base, which is supposed to be introduced in 2005). - Mobile telephone bandwidth was added to tangible assets in the 1995 base accounts. - Exceptions to scope and other limitations are explained in documents accompanying the national accounts. ### 2.3 Classification/sectorization - Classification of transactions and flows is consistent with the 1993 SNA. - Basic statistical production unit is the UHP (see 2.2.1); UHPs are grouped by industry where each industry produces a single product, enabling a square input-output table with industries and products cross-matched. - Classifications in use: - The French Activity Classification (NAF) is an adaptation of NACE. - The French product classification (CPF) corresponds exactly to the CPA. - France has an Economic Summary Classification (NES) with three levels (16, 36 and 114 items), used to compile accounts and disseminate activity and employment indicators. - Household consumption is presented under three classifications: - A product classification with 304 items based on NES levels; - A functional classification with some 40 items, the Classification of Individual Consumption by Purpose (COICOP); - A classification by durability with four items. - The breakdown of government expenditure using the Classification of Functions of Government (COFOG) is being prepared and should be published by the end of 2003. - On January 1, 2003, the NAF rev. 1 and the CPF rev. 1 replaced the NAF and CPF used since 1993; they will be used in surveys starting this year and in the 2000 base accounts. - Exceptions to classification and sectorization rules are explained in the documents accompanying the national accounts. ### 2.4 Basis for recording - Valuation: - All market output is valued at basic prices. - Output for own-use is generally valued at equivalent market prices; nonmarket output is valued using production costs. - Sales taxes, excise taxes, and nondeductible VAT are included in the valuation of intermediate consumption; valuation of final uses excludes deductible VAT. - Fixed capital consumption is valued at replacement cost for all institutional sectors. - Total imports and exports are valued on a FOB basis. - Statistical information is used to estimate insurance and freight costs (CIF) and to transform CIF to FOB values. - Transactions in foreign currency are converted at the average market exchange rate on the date when they take place. - Transfer prices are not detected. - Recording basis: - Transactions and flows of market statistical units are recorded on an accrual basis. - Work-in-progress is recorded when it takes place, except for crops and livestock (this exception has minimal effect on annual accounts and only slight impact on quarterly accounts). - Government operations recording: - Social security contributions and personal income tax are recorded for the amount owed when it becomes due. Personal income tax is not withheld at source and tax liabilities accrued in the year n are not due or recorded until the year n + 1. - VAT, oil taxes, subsidies on products, and current expenditure are recorded on an accrual basis in theory, but in practice they are recorded on a time-adjusted cash basis. - Gross fixed capital formation (GFCF) and other revenue categories, such as corporate income tax, are recorded on a cash basis. - Local government and social security accounts have been recorded on an accrual basis since 1997. The central government accounts will switch in 2005. - Value added for the general government sector includes an estimate of fixed capital consumption at replacement cost. - Grossing/netting: - Transactions between local units of an enterprise are recorded on a gross basis for a few rare products only. - Central government operations, especially redistribution operations, are always recorded on a gross basis. *Source: _cr03339 - 2. Methodological soundness* ### 3.1 Source data ### _cr03339 - 3.1 Source data ### Inventory of businesses - Complete inventory covers corporations, sole proprietorships, the self-employed, government agencies, local governments, and associations. - For each legal entity the inventory provides: the name or company name, full address of the head office and of its local establishments, legal status, an individual identification number and one for each of its local establishments (linked to that of the parent entity), turnover bracket, number of employees, and the principal industry sector according to the most disaggregated French industrial classification (NAF, which lists almost 700 industries). - INSEE updates the inventory every day using information provided by the tax authorities. - All data are entered into the SIRENE computerized business directory (managed by INSEE) and are publicly accessible via telephone, Minitel, or Internet. - The inventory is described as very complete; even the self-employed register to benefit from VAT refund. ### Surveys of enterprises — design, coverage, and use - Sampling plans and estimation procedures use recognized principles: random sampling, imputation for nonresponse, and extrapolation to the statistical universe. - Questionnaires are detailed enough to calculate most national accounts aggregates, except intermediate consumption by product (most surveys request limited breakdowns such as energy, other goods, and services). Subcomponents: - (a) The annual enterprise (EAE) and industry (EAB) surveys - EAE target all enterprises (except financial services and general government); agricultural sector findings not used in national accounts. - Two questionnaires: comprehensive for large enterprises and simplified for smaller companies. - EAE coverage: all enterprises with 20 or more employees and a sample of those with fewer than 20 employees; lower threshold in food-processing and transportation sectors. - Overall coverage as a percentage of the value added in the industry is at least 80 percent. - EAB (annual industry surveys) measure output at a highly disaggregated level; enterprises with multiple product lines receive multiple EAB questionnaires. - EAB and EAE are conducted simultaneously, with same coverage criteria; EAB helps validate the intermediate enterprise system (SIE) and serves as annual benchmark for PPI-related surveys. - (b) Subannual surveys for the industrial production index - Conducted by the SSM or authorized industry associations under DSE supervision. - Coverage: all enterprises with 50 or more employees and a sample of smaller enterprises (mostly in food-processing). - Industrial branches surveyed account for approximately 86 percent of industrial value added: monthly figures, 74 percent, and quarterly figures converted econometrically into monthly equivalents, 12 percent. - Goal: for each industry cover at least 70 percent of value added, so overall coverage is in the order of 60 percent (0.86 x 0.70). - Information varies by industry: volume of output/deliveries, billing net of taxes and unfilled orders, or number of hours worked for construction and long-cycle manufacturing. - Indices are Laspeyres indices of volumes produced (1995 = 100). - For highly heterogeneous products (approximately 35 percent of total value added), volume indices are obtained by deflating billings by an industrial sales price index. - Industrial production indices for approximately 60 branches are published in the Bulletin mensuel de statistique as unadjusted data and data corrected for WD-SA. - Series for year n are subject to revision during the same year and up to December of year n + 1; WD-SA coefficients revised four times a year and calibrated with annual surveys. - (c) Quarterly survey of products, costs, and assets - Sample of 7,500 enterprises; redrafted in 2000. - Contains accounting information used elsewhere to estimate changes in nonfarm stocks. - Could be used to calculate quarterly changes in stocks from statistical data (rather than residuals) and to integrate with sub-annual output surveys to measure output of heterogeneous industries as sum of sales (billings) and changes in stocks. - Currently, in the PPI heterogeneous output is determined by billings on the basis of 35 percent of the value added, which excludes changes in stocks. - (d) Monthly survey of commercial establishments - Stratified sample of some 8,000 commercial establishments conducted by the BdF. - Variable tracked: turnover of each establishment, including all taxes, broken into 12 broad product categories. - Value indices calculated by the panel method (fixed sample), then deflated by consumer price indices, and corrected for WD-SA. - Findings (except automobiles) serve as indicators in quarterly accounts and provisional annual accounts. - This survey is the only current source providing a partial breakdown of household expenditure on goods. ### Data from administrative records — sources and practices - Systematic use of administrative source data is a central strength: individual identification numbers for establishments are used across taxes, customs, social security, and an industry code; all enterprises use the same general chart of accounts. Subcomponents: - (a) Tax data and the intermediate enterprise system (SIE) - For large nonfinancial enterprises (approximately 100,000 establishments), tax data are systematically checked against EAE and maintained in SUSE. - Checks: completeness, internal, temporal, and inter-source consistency; where irreconcilable differences arise, tax source data are generally kept. - For nonfinancial enterprises with fewer than 100 employees, SUSE is used to extract a representative sample of original tax data; consistency analysis and imputation performed using aggregate data by activity and enterprise type. - The two subtotals (large and small nonfinancial enterprises) are integrated to produce the SIE. - SIE: database of nonfinancial enterprises with individual accounts for the 100,000 largest enterprises and aggregate accounts for smaller enterprises; structured similarly to national accounting requirements. - The SIE is a starting point for goods and services accounts and institutional sector accounts; DCN systematically adjusts turnover and value added by activity using coefficients based on tax audits. - (b) Monthly turnover indices - Base year 1995 = 100, derived from VAT records but actually based on turnover. - Published for over 600 industries (level 4 of NAF). - Monthly comparable sample of approximately 125,000 enterprises for previous and current year to calculate year-on-year rates of change. - Procedure underestimates growth due to absence of new enterprises; correction: turnover of enterprises established during the previous year are re-entered the following year and indices re-calculated. - Revised indices are published in August of each year. - These indices are used in quarterly accounts to estimate output of services and certain household consumption and gross fixed capital formation items. - (c) Data on general government - Provided by the DGCP; retranscribed in accordance with national accounts concepts. - Exhaustive accounting data used to calculate production, intermediate consumption, gross fixed capital formation, changes in stocks, transfers, etc. - Coverage: 100 percent for the final accounts and almost 100 percent for the semi-final accounts. - For provisional annual accounts and each quarter of the current year, MINEFI’s Forecasting Directorate produces a monthly estimate of central government wages and other aggregates; intermediate consumption derived as a function of wages. - DGCP will begin producing quarterly data based on an accounting system comparable with annual data. - (d) Customs data - Compiled, validated, and corrected in value terms by DGDDI before being sent to DCE of the DCN. - Regarded as exhaustive; data are revised each month and considered definitive after 24 months. - (e) Other administrative sources - Financial enterprises: accounts drawn using BdF data (banks, other financial institutions) and supervisory authorities (insurance companies, pension funds); annual accounts usually exhaustive; quarterly figures incomplete. - Construction: most data come from administrative sources (construction permits, number and size of housing projects, dates construction begins/ends, construction time matrices for four types of homes and apartments and thirteen types of building). Construction grids are monthly for quarterly accounts and quarterly for annual accounts; to be updated by MINEQUIPE in 2003 with base-year 2000 work. - Agriculture and fisheries: complete accounts drawn up jointly by INSEE and MINAGRI using a noninstitutional approach (output in value terms derived from prices and quantities). - Housing: complete accounts prepared similarly; housing services are estimated on a “quantity x price” basis and assigned to institutional sectors with balance reverting to households. ### Household surveys and sample design - Household surveys representative for Metropolitan France (overseas departments excluded) and all socio-economic groups. - Survey of household expenditure: - Conducted every five years (1995, 2000). - Covers a sample of 18,000 households. - Used in the base year to estimate several categories of household consumption expenditure, especially services. - Housing survey: - Covers 40,000 households. - Conducted every four years up to 1996 and every five years thereafter. - Used to estimate paid and imputed rents and to produce satellite housing account. - Ongoing food survey is discontinued. ### Price indices and deflation - Available price indices for deflating current price estimates or inflating volume estimates: - Consumer price indices (CPI) - Industry selling price indices (ISPI) - Farm price indices (wholesale, intermediate consumption, products) - Wholesale food price indices - Construction cost indices (construction and public works) - Raw and other materials price indices - Unit value indices for exports and imports - Wage cost indices ### Relations between data sources, sectors, and approaches - The source describes relationships between institutional sectors and sets of statistical units or activities (nonfinancial enterprises, financial enterprises, general government, households, NPISHs, agriculture, housing) and notes where units have complete accounting, supplemented accounting, or noninstitutional approaches. - Housing services estimation: estimated on a “quantity x price” basis; production, value added, and counterparts assigned to institutional sectors with the balance reverting to households. ### Key findings, data strengths, and limitations - For economic activities as a whole: - Coverage of output is practically exhaustive for the annual accounts and very high for quarterly GDP. - In the definitive accounts, coverage using tax and survey data is almost 100 percent; only part of the NPISH sector escapes measurement. - Estimate for total intermediate consumption is very reliable, implying total value added estimate is also accurate. - By industry: - Output estimates are reliable at a highly disaggregated level. - Value added estimates by activity are less reliable due to lack of detailed intermediate consumption by product, though the deficiency is not as serious as it seems. - Fiscal statistics: - Exhaustive annual fiscal statistics are available with a reasonable lag; quarterly data are far less complete but INSEE is addressing this with DGCP and BdF assistance. - Sub-annual indicators: - Monthly and quarterly indicators are sufficiently detailed, reliable, and current to allow correct calculation of quarterly GDP using a production approach. - Good sub-annual indicators exist for all demand categories, including a recently revamped quarterly survey of stocks. ### Data consistency, classification, valuation, and timing - Source data largely consistent with national accounts concepts; national accountants make adjustments where necessary (example: measurement of housing construction as work in progress). - Government revenue and expenditure usually recorded on a cash basis in public accounts, but available information is sufficient to record on an accrual basis or a good approximation. - Financial year alignment: - No legal obligation for enterprise financial year to match calendar year, but financial enterprises and approximately 70 percent of nonfinancial enterprises adopt the calendar year, reducing inconsistencies. - No routine adjustment is made for enterprises with different financial years because studies show such adjustments would be too imprecise. ### Timeliness and data lags - Major source data for annual accounts are obtained within a reasonable period. - For quarterly accounts, survey data are obtained with a reasonable lag. - Specific timing details: - Turnover indices obtained 70 days after the reference month; not available for preliminary quarterly estimates but available 20 days prior to dissemination of revised estimates (in 90 days as of first quarter 2003). - Exhaustive annual fiscal statistics available with a one-month lag for the government account and a six to 12-month lag for other units. _Italic: Source: _cr03339 - 3.1 Source data (PDF chapter/section)_ ### 3.2 Statistical techniques ### 3.2 Statistical techniques ### 3.2.1 Data compilation employs sound statistical techniques - Accounts are compiled mainly on the basis of an annual input-output table of 118 industry sectors by 118 products, (480 products for the supply and use table). The quarterly accounts use a smaller table (approximately 40 industries x 40 products). - Principal indicators used in the quarterly accounts (source / periodicity / lag in days): - Foreign trade - Exports / imports of goods — DGDDI / Monthly / 43 - Unit value indices — INSEE / Monthly / 45 - Balance of payments — BdF / Monthly / 42 - Consumption - Commercial sector survey — BdF / Monthly / 45 - Expenditure on health care — CNAM / Monthly / 45 - Licenses, private vehicles — MINEQUIP / Monthly / 15 - Turnover indices – services — INSEE / Monthly / 75 - Consumer price indices — INSEE / Monthly / 25 - General government expenditure — Forecasting Directorate MINEFI / Monthly / 80 - Household expenditure on gas, electricity, and rail transportation — Government enterprises / Quarterly / 40 - Production - Industrial Production Index — INSEE / Monthly / 42 - Industrial Production Index — INSEE / quarterly / 125 - Turnover indices – industry — INSEE / Monthly / 75 - Industrial selling price indices — INSEE / Monthly / 30 - Agriculture (values, volumes, and prices) — MINAGRI / monthly / 45 - Road transport-merchandise, volume — MINEQUIP / Monthly / 35 - Road transport-merchandise, prices — MINEQUIP / Monthly / 100 - Wage indices, services to enterprises — Industry association / Monthly / 30 - Production, refined oil products — Industry association / Monthly / 50 - Gross fixed capital formation - Licenses, commercial vehicles — MINEQUIP / Monthly / 30 - Turnover indices – wholesalers in mechanical and electronic equipment — INSEE / Monthly / 75 - Construction starts (number and size) — MINEQUIP / Monthly / 25 - Capital repairs price index — Industry association / Quarterly / 90 - Number of construction permits — MINEQUIP / Monthly / 25 - Housing stock — MINEQUIP / Monthly / 60 - Market price index, construction — Industry association / Monthly / 90 - Construction cost index — INSEE / Monthly / 90 - Public works carried out — Industry association / Monthly / 45 - Construction and Public Works cost index — Industry association / Monthly / 90 - Footnote on nomenclature: "The summary nomenclature has 114 items, to which three additional industry sectors have to be added for nonmarket education, health, and social action, plus the fictitious branch of the financial intermediation services indirectly measured (FISIM), i.e., 118 in all." - The term "base of the accounts" refers to the whole set of accounts (benchmark year). 1995 is the current base year; accounts for that year and nearby years were systematically analyzed by level. The next base will be the 2000 base; results expected to be published in 2005. ### Production approach — general considerations and specific items - Intermediate consumption estimation: - Limited use of fixed ratios older than five years. - EAE and tax data generally report intermediate consumption as a total or split into goods/services; detailed survey data are limited. - Supply and use tables constructed for 480 products; breakdown of intermediate consumption by product is likely reasonably accurate though some product allocations remain difficult when products have several uses. - Aggregate intermediate consumption estimate is “undoubtedly reliable,” but value added by activity may be “a little dubious” due to imperfect stock change data. - Recommendation: conduct studies of specific industries where technical ratios are inadequate rather than rapidly imposing new broad surveys. - Owner-occupied housing: - Housing services output assessed on the basis of rent for a comparable dwelling. - Estimates rely on an econometric model with variables: number of rooms; average area of rooms; year construction completed; variables for comfort and surroundings. - Coefficients calculated using information from housing surveys. - Work-in-progress and construction: - Construction and production of large facilities are treated as work in progress; cultivated land, forestry resources, fish farm stocks, and livestock for human consumption are not (although upcoming base change will measure natural increase of forests as work-in-progress). - INSEE follows 1993 SNA and uses the housing (or building) equivalents method, spreading out value of constructed buildings with construction time matrices (matrices in use date from early 1990s; to be updated by MINEQUIP during the year). - Demand-side estimation (GFCF for new housing) given precedence, using construction time matrices in preparing supply and use tables, and adjusted for informal sector construction and tax evasion. - Stocks evaluation: - Adjusted for holding gains/losses (except farm products and new housing, where changes are directly observed). - Adjustment made using accounting sources at the summary 118-product classification level and checked to approximate residual changes in stocks from the input-output table. - Consumption of fixed capital: - Estimated using the permanent inventory method. - Base for recording: - Various adjustments to ensure adequate recording include work-in-progress, crops and livestock, and government operations (described in 2.4.2). - Volume estimates (production approach): - Price deflators/indices established at a disaggregated (118-item) level consistent with concept, coverage, and reference period. - Volume of taxes/subsidies on products = volume of transactions × tax rate of the base year. - Volume of commercial markups = volume of sales × markups of the base year. - All items in the supply and use tables by product are deflated, except intermediate consumption, which is obtained as the balance of other deflated items. - Volume of value added = volume of output − volume of intermediate consumption (intermediate consumption determined as a residual). - Procedure akin to double deflation. Changes in volume of GDP presented with chained annual indices at previous-year prices. ### Expenditure approach — general considerations - Major points: - All major expenditure categories estimated independently, except changes in stocks (established by product for 118 items, largely as residuals of supply and use tables; exceptions: farm products and new housing directly observed). - Household final consumption expenditures estimated and deflated for approximately 300 items, published according to COICOP (± 40 items). - Final consumption expenditures of general government exclude residual sales. - A breakdown of general government consumption according to COFOG will be published shortly. - GFCF is calculated by branch and type of asset. - Production of software for GFCF on own account recorded since the change to base year 1995; method will be modified in the new base pursuant to the Eurostat ruling. - Expenditure on goods regarded as stores of value is capitalized. - Construction methods reference administrative source data (see 3.1.1). - Very little use of fixed ratios over five years old for total estimates in each major expenditure category. ### Foreign trade - Future revisions of customs data are projected using an econometric model to improve supply and use tables and minimize subsequent revisions to the accounts (best-practice example). - The Balance of Payments Directorate (DBDP) of the BdF makes a similar adjustment but only to the most recent month, creating an artificial difference between the two accounts (see 4.3.3). - Deflation and indices: - Foreign trade in goods is deflated using unit value indices (UVI). - Quarterly accounts: UVI applied systematically; UVI aggregated into 25 series, then seasonally adjusted. - Annual accounts: UVI sometimes less used as input-output table balancing proceeds; UVI obtained by aggregating monthly UVI and deflation performed at the most disaggregated level (200 series) and by zone (EMU, the EU excluding EMU, rest of the OECD, and rest of the world). - UVI construction: monthly elementary unit values at most disaggregated classification level, aggregated using a Paasche formula, then chained; about 200 UVI at the most disaggregated level for imports and exports. - INSEE started preparing explicit export indices for all branches of industry in 2001, expected completion in 2006; explicit indices for imports planned by 2004 (EU regulation on explicit price indices for imports of industrial products due end-2003). - Ideal transition: replace UVI with explicit export and import indices as soon as possible, and at latest by the new accounting base adoption; alternatives include producing a limited number of explicit indices for key products. - Foreign trade in services: - Balance of payments and national accounts differ in treatment of CIF ratios for imports of goods (see 4.3.3). - Expenditure by residents abroad and nonresidents in France correctly processed in foreign trade. - Lack of breakdown for travel-related outlays into private and business trips causes the whole territorial correction to be reflected in household consumption expenditure. - Volume estimates (trade and government): - Specific detailed price indices or deflators used to deflate GDP components at a fairly disaggregated level. - Implicit household consumption deflator moves in line with the CPI. - Final government consumption expenditures deflated by production costs; payroll deflated by a civil service wage index. - Volume of changes in inventories calculated using specific methods for each stock type (producer/user inventories). - Annual and monthly deflation of trade in services conducted at a highly aggregated level and not by zone (unlike trade in goods). - Improvement possible by using the balance of payments file to deflate trade in services at a more disaggregated level and by zone. ### Specific techniques used in the quarterly accounts - Two-stage benchmarking method (Bournay-Laroque — Annales de l'INSEE 36, 1979): - Stage 1 (benchmarking): estimate an econometric relationship at the annual level between the annual series and the selected subannual indicator to produce a subannual series whose annual total differs from the annual series by the regression residual. - Stage 2 (calibration): use a smoothing procedure to distribute the residual among the months or quarters. - Treatment of subannual series: - Quarterly accounts compiled using non-seasonally adjusted source data to produce gross series. - Gross series adjusted for number of working days using an in-house econometric method, then seasonally adjusted using X11 software; X12 adoption expected for 2003. - Although data-producing departments may use different WD-SA methods (X11ARIMA, TRAMO/SEATS, or in-house approximations to X12), the quarterly accounts treat all sub-annual series uniformly by redoing WD-SA calculations centrally. - Most indicators are monthly (see list in 3.1.3); quarterly indicators are converted to monthly using an econometric technique to enable uniform monthly benchmarking, seasonal adjustment, and deflation; quarterly figures obtained by summing monthly figures. - Quarterly volume estimates: - Quarterly accounts are expressed in volumes at 1995 prices (not previous year’s prices like annual accounts) to maintain additivity of accounts and aggregates. - Growth rates calculated from four quarters at 1995 prices are very close to annual rates at previous year’s prices, with differences rarely exceeding 0.1 or 0.2 of a percentage point. - The practice of expressing quarterly series in base year 1995 prices and extrapolating backwards (as far as 1978) to obtain rebased aggregates is followed to maintain additivity, but 1993 SNA does not recommend this practice (see paragraphs 16.33-16.41 and Table 16.1). Using a recent reference year for prices can distort older growth rates; differences between quarterly and annual growth rates widen the further back one goes. ### 3.2.2 Other statistical procedures employ sound statistical techniques - Source data are compatible with national accounts concepts and classifications. - Specific methods ensure exhaustive measurement of GDP with systematic adjustments for: - Production of undeclared goods and services (e.g., undeclared employment). - Understatements of earnings to evade taxes and social security contributions. *Source: _cr03339 - 3.2 Statistical techniques* ### 3.3 Assessment and validation of source data ### 3.3 Assessment and validation of source data ### Assessment of source data (censuses, sample surveys, administrative records) - Source data used by national accountants are routinely assessed for coverage, sample error, response error, and non-sampling error. - Responsibility for source-data quality lies with the producing units (especially the DSE); national accountants specify needs in the CNIS framework and via direct communication with data producers. - Both national accountants and data producers prioritize consistency across source datasets; when problems arise (e.g., restructurings of large enterprises), individual sources (EAE, EAB, SIE, PPI) are compared to identify and resolve issues. - The SIE permits a computerized individual check for the 100,000 largest enterprises, enabling the elimination of inconsistencies and correction of errors at the micro level for any relatively important matter. ### Examples of validation practices - Merchandise trade: future revisions are projected to minimize subsequent revisions of GDP (see 4.3.3). - Cross-checking: accounting sources are checked against production data for industrial enterprises. - Product supply and use tables established at a disaggregated level highlight discrepancies that are then analyzed and eliminated. - Value added rates by industry obtained from company accounts are compared with those resulting from approximations of output via product supply and use tables. ### Revision studies and use of micro information - Abundant micro-level information supports cross-checking, reducing the need for subsequent revisions; correction of source data often substitutes for purely statistical revision analysis. - The practice of projecting revisions of exports and imports of goods uses an econometric model; revision studies are conducted “upstream” of account compilation as well as “downstream.” - Quarterly-account revision analysis is performed to ensure revisions are not correlated with the business cycle or biased. The DCN published an article showing that national accounts revisions are neither correlated with the business cycle nor biased: - Gros, Emmanuel, and Prost, Corinne, “Les révisions annuelles des comptes nationaux sont-elles cycliques?” Note de conjoncture, INSEE, December 2002. - For annual accounts, long-range analysis has been undertaken using the former base and will continue using the current base when longer-period data are available. *Source: 3.3 Assessment and validation of source data (document text)* ### 5.2 Metadata accessibility ### 5.2 Metadata accessibility ### 5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques - Documentation in summary form is available on the website and on CD-ROM. - Detailed information is generally only available internally in the form of Notes de base, some of which are released upon request. - The Note de base on the general government account has been widely distributed. - Others will be made available to the public via the Internet toward the end of 2003 (see 5.1.2). - Complete documentation on the sources and methods used for the annual accounts is compiled whenever there is a change in the base. - The existence of the SNA manual eased the burden of preparing this information considerably in the case of the 1995 base; only one document on allocation operations was produced. - A summary description of the methodology underlying the quarterly accounts is available on the Internet. - A more complete version is being prepared for publication in September 2003 (will be also in English). ### 5.2.2 Levels of detail adapted to user needs - Metadata are available with varying degrees of detail and technical complexity, depending on user needs. ### 5.3 Assistance to users ### 5.3.1 Contact person for each subject field - Publications do not mention the person or unit producing the statistics or the name of a contact person in the unit concerned; they refer to INSEE, with a central telephone number generally provided. - On the INSEE website a central number is also provided. - The national accounts CD-ROM gives the e-mail address of a contact person in the DCN as well as contact information regarding INSEE Info Services. - INSEE centralizes user services: - INSEE prefers to work through media specialists in its press office to communicate with the press rather than involve data-producing staff directly. - INSEE Info Services handles other members of the public across visits, telephone, letters, or e-mail. - User audiences distinguished: (1) the general public (handled largely through the press office), (2) public institutions (central government, Parliament, other levels of government, research institutes, and universities), and (3) “participants in economic debate” (all other users). - The “Specific Services and Large Accounts” division handles complex requests from big users. - The INSEE website does not inform the public about the availability of these services and does not contain an organizational chart. - INSEE evaluated INSEE Info Services (especially phone services) and decided to strengthen it by creating INSEE Contact: - INSEE Contact is designed to answer all pertinent questions called in to a single telephone number either directly or by forwarding the call to the data-producing unit. - Because of this design, there are no plans, for the time being, to identify producing units or persons in charge at the foot of published statistical tables. ### 5.3.2 Catalogues of publications, documents, and other services - Every year, INSEE updates a catalogue of the publications, documents, and other services it provides, indicating any charges involved and how to order. - The catalogue is widely available at INSEE’s head office or regional bureaus. - The catalogue is also posted on the INSEE website. *Source: _cr03339 - 5.2 Metadata accessibility* ### 1.2 Transparency ### 1.2 Transparency ### Availability of terms and conditions - Information on provisions governing the collection, processing, and dissemination of data are available to the public on INSEE and CNIS websites as well as in their publications. - The consumer price index forms part of the public surveys program published in the Journal Officiel. ### Internal governmental access prior to release - The Office of the Minister of Economy, Finance, and Industry is informed of the latest CPI data at approximately 6:00 p.m. on the eve of their publication, which takes place at 8:45 a.m. the following day. - The public is informed of this practice through the metadata posted on the IMF’s DSBB. - The public is not informed of this access in the French dissemination channels. ### Identification of statistical products - INSEE is clearly identified as the source of the data. ### Advance notice of major methodological changes - The public is usually notified of major changes in the CPI methodology when they are implemented. - There is no advance notice of major changes before the publication of the first monthly index to which they apply. ### Ethical standards - All civil servants and government agents are bound by the obligation of professional secrecy; by the special legislation on computerized data; and by the code of ethics that governs relations between civil servants and the private sector. - There are penalty provisions for violations of the code of ethics. - Ethical rules relating to professional secrecy and statistical secrecy do not seem to be presented systematically to staff in a document or individual letters explaining the rules, except for new arrivals. - Survey questionnaires all mention statistical secrecy and staff who work on individual data are thus very conscious of this requirement. --- ### 2. Methodological soundness ### Concepts and definitions - For the CPI, there are currently no internationally accepted standards comparable to national accounts; the European Union has adopted regulations governing the HICP. - Methodological soundness is assessed on the basis of principles set forth in the System of National Accounts 1993 (1993 SNA), the European System of Accounts (ESA 95), the EU’s harmonized CPI (HICP) regulations, and the Consumer Price Index Manual (CPI Manual) drafted by the ISWGPS (draft chapters posted on the ILO website). - In this assessment, all references to SNA 93 include ESA 95. ### Scope - The CPI covers a selection of aggregates corresponding to monetary final consumption of market goods by households. - The CPI covers approximately 75 percent of official household consumption in the national territory. - Excluded from coverage are: - consumption by households of market goods they produce for their own use (0.4%) - imputed rents (10.2%) - imputed financial services (0.2%) - consumption of nonmarket goods (14.2%) - illicit products (drugs, prostitution, etc.) - The CPI targets all resident households (including people living in institutions and communities) and is based on the notion of domestic consumption (which includes that of non-residents in the national territory). - Current CPI coverage is 95 percent of its theoretical scope (base year 1998 in 2002). Gaps in coverage are: - 04.1.1 (part of) actual rents (0.2%) - 06.3.1 hospital services (1.9%) - 09.4.3 lotteries and gambling (0.8%) - 12.4.1 (part of) social action (0.5%) - 12.5.1 life insurance (1.2%) - 12.7 community activities (0.2%) - Imputed rents are noted as a possible exception among acceptable omissions; INSEE publishes an acquisition price index of owner-occupied housing based on administrative sources. - Discrepancies vis-à-vis ESA 95: exclusion of imputed expenditure (especially imputed rents) and recording prices of services for the month in which consumption at the observed price may start rather than the month in which delivery is completed; these are aligned with EU HICP requirements. - Meetings with Eurostat are held several times a year to discuss HICP concept application. ### Classification / sectorization - The product classification system used is the international COICOP, as recommended by the SNA 93 and ESA 95. - The CPI covers all resident households, including collective households, and does not exclude any level of income. ### Basis for recording - Prices used are generally market prices and include goods sold at special sales prices or with large discounts; discounts reserved for a limited group of households are not taken into account unless major. - Secondhand goods are not excluded on principle (second hand car prices are monitored). - Prices are collected throughout the month so the CPI represents an average of the price level for the reference month; for centrally collected utility rates, INSEE notes the date of price changes and calculates an average for the month as a whole. - On “gross” vs “net”: - The French CPI reflects the total price and ignores subsidies (education, health), while the HICP reflects only the non-subsidized portion. INSEE might consider adopting the net price approach to align with HICP. - For insurance, the French CPI follows HICP principles: the “net” household expenditure approach is adopted (CPI reflects the service portion of the total premium). European regulations require the gross premium as an approximation to the net premium when net detection is impossible. --- ### 3. Accuracy and reliability ### Source data collection programs - Weights should reflect the real distribution of household expenditures; two main sources: household budget survey and national accounts household consumption weights. - The index weights are updated (and not revised) annually. - National accounts are the principal source; supply and use tables with 472 items enable calculation of weights for 305 CPI items. Estimates corresponding to year (n-2) are updated to base month December n-1 with the help of CPI price trends. - For approximately 25 percent of the CPI weighting, a volume update is performed whenever major changes in volume arise using complementary sources (sample groups of retailers and wholesalers, occupational data, data from social security agencies, and household surveys such as the five-yearly family budget survey). - Weights of varieties within items are updated less regularly (between 10 and 20 percent are updated each year). - The weights are updated every year; the CPI is a chained Laspeyres-type index. ### Price survey design and sample - Approximately two-thirds of all prices are ascertained by interviewers visiting retail outlets. - Sampling stages: 1. A probability sample of 96 agglomerations after stratification by region (NUTS-1) and size of urban unit. Towns/agglomerations divided into four population groups; lowest tranche between 2,000 and 20,000. - The four overseas departments (DOM) are covered; their representativeness is strengthened to enable a CPI for each of them. The weight of the DOMs is less than two percent of France as a whole. - Corsica is not covered by the CPI. - The sample and agglomeration weights are reviewed every 10–15 years when a new index base is introduced and in light of recent demographic census data. 2. List of varieties for 305 items is drawn up on expert opinion criteria: representativeness, homogeneity, regular availability and sales importance. - The number of price quotations depends on associated weight and dispersion of price trends. - Elementary aggregates combine varieties with agglomerations; quotations are distributed among outlets based on an 11-category sales method typology. - Physical selection of the 27,000 sales outlets is carried out by regional Price Offices; selection of products at outlets is made by interviewers following precise instructions. - The product and sales outlet sample is updated each year during index chaining operations. - There are 1,090 types of product. To establish the price for each type, best-selling articles are selected; approximately 160,000 basic prices, covering RV and FP, are collected at retail outlets each month. - Price ratios are aggregated into 28,000 basic indices. - INSEE staff obtain data not covered by the traditional survey from centralized sources; centralized surveys collect approximately 30,000 prices each month (utilities, transportation, automobiles, medicines, etc.). - For housing services, rents of approximately 8,000 dwellings are tracked each quarter; dwellings include free-market and subsidized government sector units. - Housing units are selected by random selection from the housing census file and a recent housing file, distributed by stratum with 32 strata. - One eighth of the housing panel is renewed each quarter, ensuring full rotation in two years; 87.5 percent of the sample is tracked from one quarter to another. - Primary rent data are collected quarterly; monthly estimates are obtained using a projection model (procedure established in 1993). - With rents included, the CPI is calculated on the basis of approximately 200,000 monthly prices. Comparable country ranges: 130,000 in the United Kingdom and 300,000 in Italy. ### Source data consistency and timeliness - Source data are consistent with CPI definitions, scope, classifications, recording times, and reference periods. Prices are collected throughout the month permitting an average for the period; trade promotions and sales are reflected; utility rates are assessed taking into account the date they take effect. - Weights used in January of year t are principally based on assessments of consumption totals established for year t-2. For approximately one quarter of the weights, specific volume estimates at December t-1 are established on the basis of appropriate specific sources. - Price data are generally timely; some delays may occur when tracking relies on chains of government information (e.g., medical care from social security records, household garbage collection from tax records). - INSEE disseminates a provisional index on approximately 10 days after the reference period; the definitive index appears on the 25th of the month following the reference month, a little longer than in comparable countries. *Source: _cr03339 - 1.2 Transparency* ### 3.2 Statistical techniques ### 3.2 Statistical techniques ### 3.2.1 Data compilation employs sound statistical techniques - Degree of product disaggregation: data published monthly for the 161 groups at the five-figure level of the classification used (the four-figure level being the four-figure COICOP subindices used for the dissemination of the European HICP). - Item-level data: established at the 305 item level are published once a year, as an annual average; not published monthly because not considered reliable enough. - INSEE produces in-house data by item and variety for analysis of detailed monthly trends. - CPI weights: determined on the basis of households’ actual final consumption of market goods. - HICP weights: calculated on the basis of household consumption expenditures. - Valuation: value of goods and services represents the economic cost for the consumer at the moment of purchase, regardless of method of payment. - Calculation of basic aggregate indices: based on either the average prices ratio (homogeneous varieties) or the geometric mean (heterogeneous varieties). - Aggregation: performed every year in December using a chained Laspeyres index; weights reviewed every year and adjusted in accordance with changes in prices to bring them into line with the price reference period, which is December of the preceding year. - Imputation for unobserved monthly prices: imputed from trends for identical products surveyed at similar outlets in the same or other regions; substitution practice applied from the first month lacking data; no automatic carry-forward of previously observed prices. In line with HICP regulations, prices may be imputed for at most two months; thereafter the product is replaced by another. - Missing seasonal products (other than fresh produce): prices kept constant at the level of the last observed normal (i.e., non-discounted) price. - Permanently unavailable products: replaced in sample by other products; different adjustment techniques used to reflect differences in quality. - Introduction of new products: - If meeting specifications established at the start of the year for a CPI variety: may be introduced during the year when replacing disappeared products. - If fundamentally new: introduced the following December during the annual chaining, provided they have a sufficient market share. ### 3.2.2 Other statistical procedures employ sound statistical techniques - Seasonally adjusted overall index: calculated and published every month. - Indices by product: not seasonally adjusted. - Underlying seasonally adjusted index: INSEE publishes every month an “underlying seasonally adjusted” index for the entire set of products, which excludes fresh products and certain other products of a seasonal or volatile nature. *Source: _cr03339 - 3.2 Statistical techniques* --- ### 3.3 Assessment and validation of source data ### 3.3.1 Routine assessment of source data (coverage, sample error, response error, nonsampling error) - Sample error analysis: carried out during sample optimization efforts in 1992 and 1997. - Overall margin of error (two standard deviations): less than 0.1 percent for the annual change in the overall CPI. Example given: a change of 2.1 percent means it is between 2 percent and 2.2 percent. France’s CPI observes the 0.1 percent margin required for the HICP. - Measurement errors: impact not assessed; INSEE considers it unlikely to be significant because price collections are done directly by trained INSEE staff. - Imputation disclosure: information on extent of imputation methods provided each month. Statistical effect of imputation methods assessed as less than 0.1 percent every month. - Monitoring of extreme price changes: very marked and very slight changes systematically monitored; collected data and calculation methods checked; detected errors corrected. No automatic procedure for correcting extreme values is employed. - Ex post spot checks: collected data subject to ex post spot checks comparing interviewer-collected data with reality (a Price Office staff member does the same round as the interviewer on the same day). - Administrative/source data assessment: routinely assessed in coordination with producing agencies (e.g., by “negotiation” with suppliers). Source data analyzed every month in conversations with statisticians in producing agencies to guarantee CPI accuracy. Analysis focuses on: - consistency of observed data over time; - plausibility of monthly price changes; - correct identification of products; - in the event of substitution, correct selection of the replacement product (and possibly the new outlet) and use of an appropriate substitution method. *Source: _cr03339 - 3.2 Statistical techniques* --- ### 3.4 Assessment and validation of intermediate data and statistical outputs ### 3.4.1 Validation of intermediate data against other information - Intermediate and overall CPI data compared against other price indicators from different sources, such as producer price indices or implicit price indices from National Accounts adjusted by implicit volume indices when reliable estimates possible. - Caution: differences in scope, date, and methodology between sources mean comparisons should be carefully examined. ### 3.4.2 Assessment and investigation of statistical discrepancies in intermediate data - Unusual index changes triggered by major sectoral fluctuations or processing of specific sources are systematically investigated; detected errors are corrected. ### 3.4.3 Investigation of discrepancies and indicators of problems in outputs - Discrepancies between indices for product “families” in the same sector are automatically analyzed using computer software. - Outliers from observation errors or erroneous processing are corrected. - In some cases, sampling or comparison methods over time are reviewed to re-establish correct CPI variations in subsequent periods. *Source: _cr03339 - 3.2 Statistical techniques* --- ### 3.5 Revision studies ### 3.5.1 Routine studies and analyses of revisions - France’s CPI dissemination practice: initially disseminated as a provisional index; official CPI (definitive index) published approximately 12 days later. - Studies and simulations: routinely undertaken to improve CPI methodology; annual chaining of indices allows many methodological improvements to be introduced swiftly. - Definitive CPI revision policy: the definitive CPI is never revised, even if a major error is detected, because numerous contracts are indexed to the CPI and backward extrapolation would trigger disputes and management problems. Some countries exceptionally revise major errors, but France does not. The text notes a risk of significant error and suggests analyzing that risk to formulate future policy. *Source: _cr03339 - 3.2 Statistical techniques* --- ### 4 Serviceability ### 4.1 Relevance - User consultation: CNIS provides a forum where producers and users discuss the CPI work program to ensure relevance and practical utility. - DPC user contact: DPC maintains constant contact with principal users (government and business statisticians and economists, professional associations, consumer associations, the media, individual users). - Formal user group: no permanently constituted group, but in-depth consensus-building is attempted for major methodological changes (e.g., new index base in 1999, base year 1998 = 100). - Annual approval: collection of data for the CPI is approved each year by CNIS as part of its annual program of mandatory statistical surveys, listed in an order published in the Journal Officiel. - User feedback: Press Office and INSEE Info-Service forward user comments and responses to the DPC for inclusion in “Frequently Asked Questions.” - Special requests: INSEE responds to requests requiring specific processing, such as calculating subaggregates or backward extrapolation of indicators after classification changes. ### 4.2 Timeliness and periodicity - Timeliness: monthly estimates of the CPI of month m are published between day 10 and 12 of month m + 1 for provisional data and around the 25th of month m+1 for final data. Lags are in line with the SDDS. - Periodicity: CPI periodicity is monthly, in line with the SDDS. ### 4.3 Consistency - Internal consistency: estimates from all classification typologies are consistent with one another. - Historical consistency: detailed series consistency over time is problematic when nomenclature changes occur (1993 and 1999). Current series by group (161 groups) and item (305 items) have been extrapolated backward to 1990. - Breaks: Any break in the statistical series is indicated in publications; a methodological paper explains reasons and how to correct them. - Monthly commentary: trends in the overall CPI and principal components are analyzed every month in a commentary accompanying the release. - Consistency with other frameworks: generally, weights and indices for CPI product “families” are consistent with national accounts statistics. ### 4.4 Revision policy and practice - Publication schedule: release schedule for CPI and other periodical INSEE statistics is released to the public in advance. Provisional estimate disseminated some 10 days after reference period, followed by definitive CPI published 12 days later. Definitive index is never revised. - Annual weights revision: process during chaining follows same schedule every year. Annual weights revisions reflected in the February publication of the January CPI; provisional January index is not disseminated—only the definitive index is published (example given: on February 25, 2003). Since definitive indices are never revised, weight changes have no impact on indices already published. - Documentation: publications and database contain documentation on revisions to CPI weights; INSEE informs users that more detailed information can be obtained directly from INSEE. - Clear identification: users are advised when data are provisional or definitive. Provisional data are only disseminated for some large categories. - Studies/publication of revisions: as noted in 3.5.1, the definitive CPI is not revised. *Source: _cr03339 - 3.2 Statistical techniques* --- ### 5 Accessibility ### 5.1 Data accessibility - Presentation: CPI statistics are well presented with tables and charts to facilitate analysis. - Dataset detail: datasets published at several levels of detail; each publication contains analysis for the current period. - Time series availability: data published at detailed level for the past 13 months; longer time series available for aggregate series only. - Dissemination media and formats: - Provisional data: four-page publication (INSEE’s Informations Rapides). - Final data: eight-page publication. - More detail: provided in INSEE’s monthly statistical bulletin. - INSEE website: contains detailed CPI data for the past 13 months; longer series provided in aggregates only; access to longer-term series or metadata described as not very user friendly. - Publication schedule: CPI publication follows a schedule published four months in advance. - Simultaneous access: data made available to all users at the same time. - Nonpublished subaggregates: nonpublished subaggregates are confidential; specific subaggregates may be calculated in response to a request by certain users. ### 5.2 Metadata accessibility - Documentation availability: metadata published in brochures and articles; certain texts available on INSEE or CNIS websites. - Internet metadata limitations: volume of information easily accessible through the Internet is somewhat limited, especially on methodology; less information on the CPI than on the PPI. A guide “Pour comprendre l’indice des prix” (“Understanding the Price Index”) last edition was in January 1999; useful but somewhat out of date and not clearly flagged on the website. Recommendation: update this document and make it available in print and on the website. - Levels of detail: metadata available in varying degrees of detail and technical complexity to meet needs of different users. - Confidentiality rules limiting dissemination: - (a) Principle of not divulging the description of the products monitored at the variety level (one level above specification of the object whose price is surveyed). Example: whether the index contains the price of Camembert cheese is not published; practice noted as restrictive and could be reconsidered. - (b) Principle of not divulging information on how prices are derived from public utility rates (rail transportation, electricity, services with charges varying by personal circumstances or moment of consumption). Calculation relies on service providers giving confidential commercial information; public disclosure of providers’ business practices not contemplated. However, publishing the methods used by the DPC to derive the price of the service included in the index is considered feasible, using fictitious numerical examples to illustrate methods. *Source: _cr03339 - 3.2 Statistical techniques* ### 5.3 Assistance to users ### 5.3 Assistance to users ### 5.3.1 Contact persons and user support - Users can obtain information via an audio response system, via Minitel, over the Internet, by calling the Information Offices of INSEE’s Regional Directorates, or by writing to INSEE’s Info Services. - The names of contact persons responsible for producing the index are published on the CNIS website and in certain published documents, but not in the statistical publications themselves. - All inquiries from journalists are routed through the press office. - In the PPI context, INSEE appoints a contact person to help respondents fill in and send necessary forms and to ease respondent workload. ### 5.3.2 Catalogues of publications and charges - INSEE updates a catalogue of its publications, documents, and other services every year, indicating any fees charged for services and describing how to order them. ### Related operational practices affecting user assistance (extracted from adjacent sections) - Publication timing requirement under EC Commission Regulation N°1165/98: indices must be published no more than 35 days after the end of the reference month; France publishes them within 30 days. - Confidentiality rules relevant to respondents: - Statistical tables based on individual data may not reflect the results of fewer than 3 enterprises. - No result is disseminated when a single enterprise or establishment accounts for more than 85 percent of that result. - Staffing and contact structure relevant to user support and respondent liaison: - PPI compiled by two units supervised by the Industrial Price Division: a 17-member team at INSEE head office in Paris and a second 12-member team in Lyon. - A network of 8 senior survey officials (half of whom are based in Paris) handles direct contact with surveyed establishments. - Procedures and user-facing systems under development: - Security measures: physical files under lock and key; periodic shredding of outdated documents; results labeled “confidential” until published; security cards to access individual computers (planned); high level of protection for the transmission network for the on-line question and answer service (project under way). - INSEE hired an external consultant to develop an on-line answering system for monthly PPI questionnaires; a pilot version demonstrated advanced stage and user-friendliness. - The system will include automatic e-mail reminders to respondents and data validation to flag extreme data and request explanations. - The new system is scheduled to come on stream in 2003. - Data collection and follow-up timing relevant to assistance and reminders: - Automatic reminder procedure for enterprises that have not replied (towards the tenth of month M+1). - Phone call reminder from the official responsible for enterprises that have not responded to the automatic reminder (as of the eighteenth day of month M+1). - Validation phone call for enterprises reporting major price variations or product changes (as of the eighth day of month M+1). - Provisional estimates prepared towards the eighteen day of month M+1; final calculations at the end of month M+1. *Source: _cr03339 - 5.3 Assistance to users* ### 1.3 Ethical standards ### 1.3 Ethical standards ### Ethical rules and staff awareness - All civil servants and government agents are bound by: - the obligation of professional secrecy; - the special legislation on computerized data; and - the code of ethics that governs relations between civil servants and the private sector. - Penalty provisions exist for violations of the code of ethics. - Ethical rules relating to professional secrecy and statistical secrecy "do not seem to be presented systematically to staff in a document or individual letters explaining the rules, except for new arrivals." - Survey questionnaires all mention statistical secrecy and staff who work on individual data are thus very conscious of this requirement. --- ### 2. Methodological soundness ### 2.1 Concepts and definitions - "The concepts and definitions used in preparing PPI follow international standards and recommendations (ESA 95, EC Regulation 1165/98 concerning short-term statistics, 1998) and Eurostat’s methodological manual on short-term statistics (April 2002 edition)." - For the purposes of this ROSC, the standards applied to the PPI are the EU legislation and methodological manual. - "In the event of discrepancy with these standards, reference is made to the new international manual (PPI Manual nearing completion, available on the IMF website, http://www.imf.org/np/sta/tegppi/index.htm.)" ### 2.2 Scope - Definition: "The PPI (also called ISP: industrial selling prices) for the domestic market measures changes in transaction prices for goods produced by industry and sold on the domestic market." - Scope corresponds to sections C, D, and E of the NACE (rev 1), with the following exceptions: 12.0, 22.1, 23.3, 29.6, 35.1 and 35.3. - Eurostat excludes activities related to uranium mining, publishing, the manufacturing and processing of nuclear material, the manufacturing of arms and ammunition, ship building, and airplane and spacecraft construction. Thus the index covers mining and energy sector activities. - "This report refers only to the PPI for the domestic industrial market (industry comprises approximately 18 percent of France’s GDP)." - INSEE also produces other producer price indices: - (a) PPIs for external markets; - (b) import price indices; - (c) business services price indices for the domestic market; - (d) producer price indices for farm products. ### 2.3 Classification / sectorization - Coverage: resident establishments (enterprises) of the productive sector, excluding the overseas departments (DOM) and Corsica. - Nomenclatures used: NES, CPF, NAF, H-MIG and ISIC (international or French transpositions). - Classification mapping presented: - International level: ISIC rev 3 - European level: NACE rev 1(500 categories) - France: NAF (700 categories) - Products: CPC, CPA, CPF (2300 sub-categories) ### 2.4 Basis for recording - 2.4.1 Market prices: "Weights and prices are assessed on the basis of the sales prices of the goods produced." - For base year 1995: operations monitored excluded all taxes and freight charges. - For base year 2000 (in effect in 2003): data monitored include non-deductible taxes and exclude VAT and freight charges. - 2.4.2 Accrual basis: "Operations are recorded when billed, regardless of the date of production..." - Inter-establishment transfers: "Sales or transfers of goods and services between establishments in the same group are excluded ... Respondents are informed of the exclusion of transfer prices during the five-yearly visit of the senior survey officials. The corresponding turnovers are likewise excluded from the weights..." - 2.4.3 Grossing/netting: discounts granted to the buyer are deducted; prices correspond to sales prices of transactions billed during the referenced month. --- ### 3. Accuracy and reliability ### 3.1 Source data - Sample frame: based on annual enterprise surveys drawing on the SIRENE computerized business directory. - Enterprise selection: - Enterprises are ranked in decreasing order of sales. - Generally, enterprises whose accumulated turnover represents 70 percent—and never less than 50 percent—of the turnover of the industry are used for the survey. - In highly concentrated sectors, higher figures may be attained. - Sampling type: "judicious" or "cutoff" sampling. - Coverage and respondents: - "The PPI covers 90 percent of industry." - "Altogether 4,350 suppliers take part in the monthly PPI survey, reporting 24,000 prices each month." - Industry associations are consulted before an industry is introduced or updated; final enterprise sample reflects association data, reorganizations, and special circumstances. - Imputation methods used when no reply is received include: - imputation of prices based on changes in the prices of similar products; - carry-forward of the last price surveyed; and - imputation of prices based on changes in the prices of other suppliers, etc. - Some industrial services cannot be monitored due to inability to identify sufficiently similar services (examples given: item 29.23.9 of NACE rev. 1 and specific measurement and control instruments). ### 3.2 Statistical techniques - Data collection: initial visit to enterprises and later by mail. - Weight and product list revision: five-yearly cycle; "20 percent of the establishments are covered every year" (renewal). - Interview process: - Senior technical surveyor (often with industry experience) conducts interviews lasting one or two hours, gathers transaction volumes by market, and defines representative transactions using Eurostat’s PRODCOM classification. - Turnover figures for the last known accounting year are collected for weight calculation and retained for approximately five years. - Publication detail: data published each month at four-figure CPF levels, often in greater detail, and at the NES levels. - Aggregation method: Laspeyres index; "The current reference period of the weights and the reference period of the prices for the index are the same." - New products: mainly incorporated during complete renewals at five-year intervals; disappearance and replacement of products fully reflected at a change of base year. - 3.2.2 Quality adjustments: - Hedonic methods used for office computers, laptops, and servers based on a catalogue of 30 to 50 detailed specifications for approximately 500 models; data provided by a subcontractor and analyzed by the PPI division. - Apart from hedonic cases above, the division does not carry out any other special price adjustments. ### 3.3 Assessment and validation of source data - Routine monitoring: - Changes in prices that could distort the index are systematically monitored; if an error is detected the anomaly is corrected. - "No automatic procedure for correcting extreme values is employed." - INSEE has not evaluated the impact of measurement errors and lacks a method for routine assessment of potential errors. - Monthly source data analysis focuses on: - consistency of observed data over time; - plausibility of monthly price changes; - correct identification of products; and - in the event of substitution, correct selection of the replacement product (and, possibly, the new outlet) and use of an appropriate substitution method. - Abnormally low or high prices are confirmed by respondents; reprocessing is studied case-by-case by the sector expert. - Electronic collection: "the on-line response site, which will be operating in 2003, will not have an interactive control system for the reply sent by the enterprise over the Internet." A data validation tool for in-house use at INSEE will include this capability. ### 3.4 Assessment and validation of intermediate data and statistical outputs - Intermediate/global PPI data are compared with other price indicators (e.g., consumer price indices, raw materials price indices) to test plausibility; examples include routine CPI–PPI comparisons for vehicle models. - Monthly prices are examined and apparently abnormal data are checked directly by phone with respondents (this occurs several hundred times per month). - Unusual variations triggered by major sectoral changes or specific processing sources are systematically scrutinized and corrected if errors are detected. - Discrepancies between indices for product “families” in the same sector are analyzed; outliers stemming from observation errors or processing errors are corrected. *Source: _cr03339 - 1.3 Ethical standards* ### 3.5 Revision studies ### _cr03339 - 3.5 Revision studies ### Revision studies (3.5) - As the monthly revision of data is a systematic procedure with little statistical impact, there is no real need to carry out special revision studies. - Reporting cadence for findings after the reference month: - On the sixteenth and eighteenth day of the month following the reference month, 60 percent of the findings have been reported. - This increases to as much as 90 percent of the findings between the 23rd and 26th of the month when the PPI is finalized for dissemination. - Conclusion: No dedicated public studies or analyses of revisions are performed (see 3.5.1 and 4.4.3). ### Serviceability — Relevance (4.1) - Monitoring and user engagement: - The CNIS provides a forum where both producers and users of statistics discuss the PPI work program, assuring that relevance and practical utility meet users’ needs. - The Industrial Sales Prices Division maintains constant contact with principal users (government and business statisticians and economists, professional associations, consumer associations, and the media). - INSEE responds to specific user requests requiring special processing, such as calculating subaggregates or backward extrapolation of indicators in the event of a change in classification. ### Serviceability — Timeliness and periodicity (4.2) - Timeliness: - The PPI is disseminated not later that the 30th of the month (m+1), in line with the SDDS and Eurostat standards (4.2.1). - Periodicity: - The PPI is published every month in line with the SDDS standards (4.2.2). ### Serviceability — Consistency (4.3) - Internal consistency: - Estimates resulting from all the classification typologies are consistent with one another (4.3.1). - Temporal consistency: - Availability of detailed series consistent over time is more problematic when a change of nomenclature takes place, especially for certain aggregate levels. - The current series were extrapolated backward to 1980 when such series were available (4.3.2). - Cross-framework consistency: - Generally, weights and indices for PPI product “families” are consistent with national accounts statistics (4.3.3). ### Serviceability — Revision policy and practice (4.4) - Transparency and scheduling: - Users are advised on the INSEE website that initial published data are preliminary and subject to revision. Revised data are disseminated in the same way and with the same level of detail (4.4.1). - The producer price index data are provisional for six months to allow inclusion of late data (4.4.1). - Practical impact of revisions: - In practice, revisions only involve the very latest observations and are fairly minor. - Given the minor nature of revisions, INSEE could consider finalizing the index earlier (for instance after three months) or informing users that revisions do not make a significant difference (4.4.1). - Identification of preliminary data: - Initial published data are clearly identified as preliminary on the INSEE website; revised data items carry an “r” alongside them (4.4.2). - Public studies of revisions: - As noted in 3.5.1, no revision studies are made public (4.4.3). ### Accessibility — Data accessibility (5.1) - Presentation and user aids: - PPI statistics are published in the Bulletin Mensuel de Statistique with clear and detailed presentation; data are disseminated on the “Indices Pro” website with tables and charts and a monthly commentary analyzing changes in the aggregate PPI and principal components (5.1.1). - Datasets are published at several levels of detail; each publication contains an analysis of changes in the current period (5.1.1). - Time-series availability: - Data are disseminated in the form of time series for the past 36 months on the “Indices Pro” website (5.1.1). - Website and access: - The “Indices Pro” website is user friendly, attractive, and offers a wide range of choices (base year for indices, graph presentation, degree of detail). The website is free to visit at: http://indicespro.insee.fr (5.1.1). - Dissemination media and formats: - Initial dissemination occurs in an eight-page publication (INSEE’s Informations Rapides); more detail follows in INSEE’s monthly statistical bulletin (5.1.2). - Recent data and the aggregate time series are published on the INSEE website; detailed time series are accessible through “Indices Pro” (5.1.2). - Release schedule and simultaneous access: - The PPI is published according to a schedule announced four months in advance; the date of publication announced one year in advance is always observed (5.1.3). - Statistics are made available to all users at the same time, on the date and at the time officially announced (5.1.4). - Access to nonpublished subaggregates: - Preferred subscribers to the “Indices Pro” website (mainly participating enterprises) can access more detailed information than the general public (5.1.5). ### Accessibility — Metadata accessibility (5.2) - Documentation availability: - Metadata on statistical series are published in brochures and articles; certain texts are available on the “Indices Pro” website or the CNIS website (5.2.1). - Detailed information can be found in the bilingual guide “Les indices de prix de vente de l’industrie et des services aux entreprises”/ “Producer price indices and business-service price indices” – INSEE Méthodes, No. 89, October 1999 (5.2.1). - Levels of detail: - Metadata are accessible at different levels of detail and technical complexity to meet different user needs. The “Indices Pro” website is currently consulted by some 20,000 outside users each month (5.2.2). ### Accessibility — Assistance to users (5.3) - Contact and assistance: - Contact persons for subject fields are publicized; users can obtain information via an audio response system, Minitel, Internet, or by directly contacting competent managers whose names and telephone numbers appear on dissemination websites, or by writing to INSEE’s Info Services (5.3.1). - The “Indices Pro” website provides names and contact details (address, telephone, fax, e-mail) of persons who can be contacted (5.3.1). - All inquiries from journalists are routed through the press office (5.3.1). - Catalogues and fees: - A catalogue of publications, documents, and other services, indicating any fees charged and how to order them, is available on request from INSEE (5.3.2). ### Data Quality Assessment — Key observations - The assessment table (Table 3) indicates: - 3.5 Revision studies: X (Practice Observed). - Note in comments: "No computerized method for assessing measurement errors." - Serviceability and accessibility elements are largely observed across the assessment matrix (various entries marked X for Relevance, Timeliness and periodicity, Consistency, Revision policies and practices, Data accessibility, Metadata accessibility, Assistance to users). *Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2003/_cr03339.pdf* ### 0.2 Resources ### 0.2 Resources ### 0.2.1 Staff, financial, and computing resources - The DGCP office responsible for collection and processing of data for 156,000 general government units in France has a staff of 22, in addition to INSEE staff. - The DGCP is joined by staff of the three departments involved in preparing the general government accounts (INSEE, DP, and BdF). - The DGCP considers this staffing sufficient to carry out its mandate in this area. - Some delays have occurred regarding the classification of expenditure by function, and these delays are being shortened. - Budgetary constraints have prevented updating of computer equipment at the DGCP. - Given the large number of general government units, the DGCP relies on outside data processing services to store its databases. - Because of data volume, the DGCP transmits data to INSEE and to the SESOF via messaging or on CD-ROM. ### 0.2.2 Measures to ensure efficient use of resources - Preparation of government statistics follows the same measures as national accounts: work programs and assessment of different categories of resources allocated to statistical operations. - In the early 1990s the DGCP embarked on a policy of streamlining computerized data collection to avoid duplication and minimize costs (see also the reform on government finance under 0.3.1). ### 0.3 Quality awareness #### 0.3.1 Processes to focus on quality - The DGCP has undertaken a reform of government finance and government accounting driven by the organic law of August 1, 2001 on the budget law, conducive to an in-depth reform of the government financial and accounting system scheduled to be completed in 2006. - The provisions of the organic law consolidate work done by the Treasury since 1999 to enhance readability of government accounts and meet information needs of the public, Parliament, and the Cour des comptes. - EU integration fosters harmonization of government accounting systems with each other and with international standards. - The DGCP’s contribution to this reform is incorporated into a multiyear program of actions that came on stream in 2001 and envisages, inter alia, increased transparency in government accounts. - Increased transparency will be achieved by putting into place genuine accrual basis accounting in the course of implementing the accounting component of the organic law of August 1, 2001 and through efforts to modernize local accounting systems. - Reform actions include designing the new national financial and accounting information system: reforming the current system, enhancing the information produced, improving quality of the accounts, and expediting the rendering of those accounts. #### 0.3.2 Processes to monitor quality of collection, processing, and dissemination - Source data for annual statistics covering the whole of general government are final accounting data signed by public accountants and validated by their supervisory ministry. - Central government accounts are remitted for verification and certification to the Cour des comptes; local governments and their agencies are remitted to the Cours régionales et départementales. - The head of the unit responsible for compiling these data in the DGCP is traditionally a national accountant on secondment from INSEE to ensure conformity with national accounts. - INSEE receives all detailed financial statements from the Centrale de bilan of the BdF, enabling monitoring of treatments used. - Collaboration among different teams (DGCP, DP, BdF, INSEE) enhances quality; INSEE conducts quality control checks (search for explanations of changes, validation of accounting treatments, and classification). - The Cour des comptes audits government accounts annually, certifying consistency, accuracy, and reliability; its annual reports have clarified differences of treatment between government accounts and national accounts based on information from national accountants at DGCP, DP, and INSEE. - Work on interpretation of ESA 95 takes place in European working groups, especially the Working Group on Financial Accounts, and in specialized task forces organized by Eurostat. #### 0.3.3 Processes to deal with quality tradeoffs and guide planning - INSEE monitors tradeoffs between quality and timeliness while responding to user needs. - The European Commission pushes for speedier publication of statistics; INSEE emphasizes that shortening publication times must not be done to the detriment of accuracy. ### 1. Integrity #### 1.1 Professionalism - Compilation of general government account statistics follows the same integrity requirements as other national accounts data. - Government finance statistics compilation follows public accounting and national accounting rules, leaving no leeway in treatment or interpretation of transactions. - Basic sources are principally accounting and administrative; statisticians report no political interference in choice of sources. - INSEE is entitled to comment on erroneous interpretation and misuse of statistics but favors a preventive and educational approach, dealing bilaterally with journalists. - Main published statistics are accompanied by commentaries (in Informations Rapides and INSEE-Première) to prevent erroneous interpretation. #### 1.2 Transparency - Policies and practices regarding general government account statistics are the same as for all national accounts; the public is informed through laws or regulations and basic methodological documents (ESA 95, Eurostat’s Manual on government deficit and debt, and “sources and methods” papers). - Availability of methodological documents is noted in publications such as the Rapport sur les comptes de la Nation and Le compte des administrations publiques (Jean-Pierre Dupuis, Système français de comptabilité nationale, Méthode Base 95, No.22-November 2001). - Internal governmental access: data are submitted to the Office of the Minister at 6:00 p.m. on the day before release. - Deficit and debt statistics prepared under the Stability and Growth Pact are reported to the Minister of Economy, Finance and Industry before being sent to the European Commission; no public announcement is made regarding these practices. - Monthly government finance and national debt accounting statements are available for use by MINEFI departments or other competent units. - Statistical products are clearly labeled as produced by INSEE. - Advance notice of major methodological changes is usually provided when changes are implemented; large changes are sometimes announced in advance (e.g., reform following the introduction of ESA 95). - Changes in public accounting system and presentation of government finance statistics are subject to public studies and debate and are widely commented upon in media and specialized fora. #### 1.3 Ethical standards - DGCP and INSEE staff are government employees subject to governmental ethical standards, statutory and regulatory provisions regarding professional secrecy and confidentiality obligations. - Specific laws exist regarding computerized data. - No mechanism is used to systematically remind staff of ethics provisions. ### 2. Methodological soundness #### 2.1 Concepts and definitions - Annual statistics of the general government account are compiled on an accrual basis following the structure, definitions, and concepts of ESA 95 and the ESA 95 Manual on government deficit and debt. - In government finance, definitions and concepts are broadly similar to GFSM 2001; divergences are well known and direct links between the two methodologies are easy to establish. - Statistics are presented in two analytical frameworks: (1) general government “supply” and “use” through the national accounts sequence and (2) a “government finance” presentation akin to the 1986 GFS methodology (revenue, expenditure, and financing). Net borrowing/lending is identical in both presentations and corresponds to that required by European regulations under the Stability and Growth Pact. - France is producing complete quarterly national accounts for general government in accordance with European regulations for nonfinancial accounts and forthcoming financial accounts. - Monthly statements on budget execution and Treasury operations (SROT) present accounting results under three headings: - The table “Eléments d’actif et de passif” (Main Assets and Liabilities Items) with a presentation similar to the balance sheet envisaged by the General Chart of Accounts. - The table of budget and financing operations using the presentation provided in the organic law of January 2, 1959 on the budget law to track flows associated with execution of budget laws and corresponding financing operations. - The table “Opérations budgétaires de l’exercice” showing flows associated with execution of the initial and supplementary budget laws, distinguishing charges to the preceding fiscal year, the current year, or the following year. - SROT presents data on government’s budgetary operations and their financing and on outstanding assets and liabilities. #### 2.2 Scope - The scope of annual statistics of the general government account is consistent with ESA 95 principles, covering institutions and operations exhaustively and corresponding to GFSM 2001 coverage. - Annual statistics cover the entire general government sector and its operations in flows and stocks, even at preliminary stage. - Nonconsolidated statistics are shown by subsector: central government (budgetary central government proper and central government agencies (ODAC)), local government (local governments proper and various local government agencies), and social security funds (social insurance schemes and bodies dependent on social security funds), defined in accordance with ESA 95 principles. - Data are available on flows between general government units and on reciprocally held financial balances to allow compilation of consolidated statistics at the aggregate level and for the three subsectors. - Presentations cover the following stocks and flows needed to reconstitute the “government finance” framework: - Table showing expenditure and revenue of the general government sector and of each subsector. - Table of the financial account for the general government sector and the central government subsector. - Table of the financial account for stocks for the general government sector (balance sheet), at opening and closing. - Account recording revaluation and other volume changes for the general government sector in the Tableau économique d’ensemble. - Debt as defined in the national accounts (for the general government sector). - Debt as defined in European regulations (for the general government sector). - The analytical framework is similar to, but not identical with, the GFSM 2001 framework. - The monthly Treasury statements (SROT) cover operations of the budgetary central government, a subsector of central government according to ESA 95. #### 2.3 Classification/sectorization - Definitional criteria for institutional sectors and subsectorization of the general government sector are consistent with ESA 95 criteria and with GFSM 2001 criteria. *Source: _cr03339 - 0.2 Resources — https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2003/_cr03339.pdf* ### 1. Central government (APUC, S.1311) ### 1. Central government (APUC, S.1311) ### Institutional structure and coverage - Subsector breakdown: - Central government proper (État) (S.13111) - Central government agencies (ODAC, S 13112) - Local government (APUL S.1313) — Local government proper (S.13131); Local government agencies (ODAL, S.13132) - Social security funds (S.13141) — Social insurance regimes (S.13141); Agencies dependent on the social security funds (ODASS, S.13142) - Compilation and consolidation: - Statistics are compiled for each subsector and then consolidated to obtain statistics for the three main subsectors (central government, local government, and social security funds) and for the general government sector. - The SROT covers central government budget operations and special Treasury accounts (l’État, sector S1311 according to the ESA 95). - Government debt presentation: - Debt is derived from an accounting logic and broken down into negotiable debt and nonnegotiable debt (mainly correspondent bank deposits, current accounts at the BdF, short-term investments, BdF lending, beneficiaries of checks written on the Treasury, and payables). ### Basis for recording - Market prices and valuation: - Stocks and flows are valued at market prices. - Transactions in foreign currency are converted into euros at the median rate between the buying and selling rates on the day of the transaction. - Stocks are valued at the rate in force at the end of the period. - Outstanding government debt is recorded at face value in accordance with EU recommendations. - Because the SROT is an accounting statement, transactions are recorded at the actually-used price or value. - Timing: accrual basis - Stocks and flows in the general government account are recorded on an accrual basis in line with ESA 95 and the ESA 95 Manual on government deficit and debt. - Amounts unlikely to be collected are not recorded, per ESA 95. - Recording practices and timing for specific items: - Taxes based on tax assessment rolls (especially taxes on household income and local taxes) are recorded at the time/date the tax roll is issued, on the basis of amounts due on that date. For household income tax for year n paid in year n+1 (not withheld at source), year n+1 is considered the linking period. - VAT (and certain taxes on farm products) includes amounts received linked to the taxable event (the month of the activity or placement on the market). - Other taxes, including corporate income tax, are recorded upon receipt. - Employee and employer social security contributions are recorded on the basis of amounts due, linked to the taxable event (the month of activity). - The generalized social contribution (CSG) and the contribution to repayment of the social debt (CRDS) of wage earners are considered amounts due linked to the month to which the wages relate; other CSG and CRDS are recorded upon receipt. - Social benefits in cash are recorded when the right is established (in principle by administrative ruling). - Benefits in kind are recorded when delivered, or in the case of social security refunds, when agreed to by social security agencies (receipt of request for the refund). - Interest is recorded on an ongoing basis as accrued interest. - Capital transfers for taxes due that are not collectable: amount posted = difference between amounts due, net of tax relief, and amounts received. - Capital transfers for contributions due that are not collectable: amount posted = difference between contributions due based on the taxable event and amounts received. - Grants are generally recorded when disbursed rather than at the time they are due. - Time of recording in the SROT follows the government chart of accounts: budgetary revenue is recorded when received by public accountants; budgetary expenditures are reflected when public accountants approve payment orders sent by authorizing officers. - Government debt data use an accounting-based recording that may differ from Maastricht Treaty or national accounts concepts. - Grossing/netting procedures - General government account statistics: revenue and expenditure compiled on a gross basis; financing on a net basis, in line with ESA 95. - Corrective operations (e.g., tax refunds) are deducted from corresponding aggregates as per ESA 95. - Outstanding debt is recorded on a gross basis pursuant to EC Regulation No. 3605/93. - In the SROT: - Individual categories of budget revenue generally recorded gross, including rebates; reimbursements included in fiscal expenditure as applicable. - Revenues are shown net of amounts withheld for local governments and the European communities. - Cash flow operations (financing) are recorded gross (debits and credits). ### Accuracy and reliability — source data and compilation - Source data collection and coverage: - The DGCP prepares a first draft of the annual general government account for the national accounts based on accounting sources for each institutional unit (central government, ODAC, local governments, social security funds). - Source data consist of income statements and balance sheets covering necessary economic stocks and flows, though adjustments may be required for ESA 95 consistency. - Local government data: - Coverage: at the semi-final statistics stage, only 0.3 percent of local government outlays cannot be covered by accounting sources; for final statistics, coverage is 100 percent. - France collects local government data exhaustively on the basis of a master list (référentiel), updated every year and providing 99.7 percent coverage. - Adjustments include a partial survey to obtain quarterly data swiftly for the 0.3 percent not covered in semi-final statistics. - DGCP central files: - Central file contains accounts and balances of all public entities with a public accountant and other agencies classified in the government sector. - DGCP centralizes electronic files for the accounts of 36,000 municipalities and 120,000 local public entities, which use a common chart of accounts. - Social security agencies: - DGCP centralizes accounts data; since 2002, some of these are electronic files, based on a new chart of accounts prepared by the interministerial committee. - Source data alignment and timeliness: - Source data are accounting records prepared in accordance with budget law provisions; bridge tables align accounting/budget items with ESA 95 categories and can be disseminated or published. - Coverage differences between budget outcomes and government finance statistics are described in a background note. - Source data are timely for annual compilation; timeliness is being shortened to obtain semi-final data on a fiscal year within three months of its ending. ### Statistical techniques, validation, and revisions - Compilation and adjustment techniques: - Cash-basis data are adjusted to assign them to the period to which they relate (especially taxes, subsidies on products, government expenditure and revenue). - Final consumption expenditures by general government are calculated by deflating production costs for tax purposes. - Volume of changes in stocks calculated using specific methods for each type of stock (producer or user). - Source data are either consistent or alignable with national accounts definitions and classifications. - Assessment and validation: - Basic sources are accounting records subject to strict auditing and verification procedures. - DGCP produces intermediate accounts from accounting sources; INSEE examines and validates these after necessary corrections ("specific corrections by INSEE"). - Corrections are documented in explanatory notes and compiled for internal use (examples include Corrections apportées à la version de la Direction Générale de la Comptabilité Publique du compte 2000 semi-définitif des administrations publiques en Base 95). - Main aggregates (operating balance, net borrowing/lending) are reconciled and published with comparable data elsewhere; budget deficit and Treasury debt are linked to national accounts via published bridge tables. - Close coordination with the BdF helps identify and adjust discrepancies between general government financing data and monetary data. - Values not corresponding to expected trends are identified and analyzed. - Revisions practice: - Revision practice for general government account statistics follows that of national accounts. - Reasons for revisions are accounted for and analyzed; long-term series are generally not subject to large changes due to extensive use of accounting data even at preliminary stages. ### Serviceability and relevance - Users and uses: - Users include INSEE, the BdF, the government, Parliament, the European Commission, the European Central Bank (ECB), financial and academic circles, and international organizations. - Statistics are used to monitor and analyze tax and fiscal policy within a long-term macroeconomic framework and to coordinate member states’ fiscal policies (including assessment of deficit and debt criteria). - General government account statistics are used as a macroeconomic framework in formulating the annual government budget and are presented in the autumn in the Rapport économique, social et financier attached to the draft budget law (the bleu budgétaire). - User feedback and suggested improvements: - Users noted that government finance statistics in France could be enhanced by a fiscally-oriented publication providing more detailed linkages between national accounts-based statistics and administrative Treasury data on budget revenues, expenditures, and financing. - Degree of detail could be improved by completing the introduction of a functional classification; the COFOG classification will be introduced by end-2003. - Compilation work is underway and 10 years should be published by 2003. - Dissemination of quarterly government finance statistics would be desirable for short-term monitoring of tax and budget policies. *Source: 1. Central government (APUC, S.1311)* ### 4.2 Timeliness and periodicity ### _cr03339 - 4.2 Timeliness and periodicity ### 4.2.1 Timeliness follows dissemination standards - Annual general government account statistics on revenue and expenditure (and also debt data) are first published at the beginning of March (première évaluation) following the reference year and then at the end of April (comptes provisoires). - A comprehensive presentation of statistics on revenue, expenditure, and financing, as well as balance sheets, is published in June on a CD-ROM.27 - The data (complete government sector, including financing) are available in accordance with the SDDS deadline of six months after the end of the reference year. - The SROT is published in the Journal Officiel 45 days after the reference month, and in the Notes de Bercy as: - Situation mensuelle des opérations du Trésor (SMOT) and - Tableau synthétique des opérations du Trésor. - The SMOT and Tableau synthétique are disseminated on the Internet 36 days after the reference month. - The SDDS requires publication within one month; France publishes revenue and expenditure data in the Situation du budget de l’État and financing data in the SROT 45 days after the reference month (thus availing itself of a flexibility option). - Central government debt data, included in the SROT, are published earlier than the SDDS requirement (of 1 quarter). ### 4.2.2 Periodicity follows dissemination standards - General government account statistics are published on an annual basis in line with the SDDS. - Work is currently being undertaken toward the publication of quarterly data. - For the budgetary central government sector, monthly financial statement figures for budget operations and financing operations are monthly and conform with the SDDS periodicity requirements. - Statistics on outstanding central government (central government subsector) debt, including outstanding guarantees, are published every month (more frequently than required by the SDDS). *Source: _cr03339 - 4.2 Timeliness and periodicity* ### 0.1 Legal and institutional environment ### 0.1 Legal and institutional environment ### 0.1.1 Responsibility for collecting, processing, and disseminating statistics - Since the transfer of authority for monetary matters to the Eurosystem and the introduction of the euro in January 1999, the European Central Bank (ECB), assisted by the national central banks (NCBs), is responsible for monetary statistics and related statistics for the euro area. - The ECB Statute and Article 5 of the Protocol on the Statute of the ESCB and of the ECB (Maastricht Treaty, November 1993) entrust preparation of monetary statistics to the ECB, assisted by euro-area NCBs. - The Banque de France (BdF) collects source data within France; the reporting population (monetary financial institutions—MFIs—and other financial institutions—OFIs—other than insurance companies and pension funds resident in France) is defined by Council Regulation (EC) No. 2533/98 of November 23, 1998. - Reporting modalities and reporter definitions are stipulated in Regulations: - ECB/2001/13 of November 22, 2001 (replacing ECB/1998/16 of December 1, 1998, effective January 1, 2003), - ECB/2002/4 of June 6, 2002, - ECB/2002/8 on the consolidated balance sheet of the MFI sector (formats/procedures stipulated in ECB Guideline ECB/2002/5 of July 30, 2002), - ECB/2001/18 of November 22, 2001 on interest rates applied by MFIs to deposits and loans vis-à-vis households and nonfinancial corporations. - In France: - BdF responsibility for collecting, processing, and disseminating monetary statistics is conferred by Article L144-1 of the Monetary and Financial Code (CMF). - Practical implementation is set by Decision No. 02-01 of the BdF Governor and Notices Nos. 02-01 and 02-03 of the Director-General of Research and International Relations. - Data collection mechanisms: - Data from credit institutions are collected via the Database of Financial Agents Doing Business in France (BAFI) managed by the General Secretariat of the Banking Commission (SGCB) as the sole point of entry. - Over 1,000 credit institutions submit quarterly (series 4000 prudential documents); 700 submit monthly (series 8000 monetary statistics documents). - Credit institutions as defined by the 1984 Banking Law (except mutual guarantee companies) are treated as MFIs by the ECB. - The Monetary Analyses and Statistics Unit (SASM) of the BdF receives BAFI data for preparing monetary statistics (Article 2 of Decision No. 02-01 of May 22, 2002). - A steering committee for the BAFI system (including SASM, SGCB, Balance of Payments Directorate) must reach consensus on any collection process change. - Nonbank reporters: - Article 2.3 (Decision No. 02-01) tasks BdF with defining reporting requirements for MFIs that do not report to the CB and implementing tailored collection mechanisms if necessary. - SASM manages collections covering about 600 “monetary UCITS” (money market funds) and slightly over 7,000 nonmonetary UCITS. - The Caisse des dépôts et consignations and the Caisse nationale d’épargne operating at postal facilities are regarded as MFIs; SASM gathers monthly data directly from them. - SASM collects monthly information on demand deposits managed by the Public Treasury and the post office (included in euro-area monetary aggregates), though these two institutions are not regarded as MFIs. - Outputs: - BdF produces statistics at the national level applying a residency criterion extended to the euro area (the “French contribution”) and forwards results to the ECB, which compiles consolidated MFI-sector statements for the euro area. - SASM also produces an analytical national statement (Bulletin mensuel de la BdF, Statistics, Table 11) and an aggregated statement for other resident MFIs (Table 12) using a national residency criterion. ### 0.1.2 Data sharing and coordination among data producing agencies - Legal basis for exchanges with the Banking Commission (CB) is Article L631.1 of the CMF, which authorizes specified entities (including the BdF, CB, Banking Commission, Insurance Supervision Commission, and others) to communicate necessary information, subject to professional secrecy. - Article 2 of Decision No. 02-01 (May 22, 2002) authorizes BdF to use the Banking Commission’s information collection mechanism (Banking Commission Instruction No. 94-09 of October 17, 1994) and the balance-of-payments compilation system, supplementing as necessary for monetary statistics; submissions remain addressed to the Secretariat General of the Banking Commission. - Regular mutual consultation exists between units responsible for monetary statistics, the financial account of national accounts, and the balance of payments—particularly during the January 2003 strengthening of the consolidated balance-sheet collection (ECB/2001/13). - Coordination mechanisms: - Modifications are worked out jointly in working parties and an online review with the Balance of Payments Directorate. - Monetary statistics, financial account, and balance of payments units are all within the Directorate General for Research and International Relations (DGEI) at BdF, facilitating coordination. - BdF participates extensively in ESCB committees, working parties, and workshops on statistical issues. - BdF participates in CNIS work to avoid duplication of data collection; the BdF Governor is an ex officio CNIS bureau member (represented by a DGEI deputy); the rapporteur for “Money, Finance, and Balance of Payments” is a senior DESM staff member. - BdF forwards its statistical work program annually to CNIS. ### 0.1.3 Confidentiality and use of respondents’ data - Confidentiality provisions for all ESCB entities are defined by Council Regulation (EC) No. 2533/98 of November 23, 1998 (Article 8, paragraph 5): collected data are to be used solely for performing ESCB missions, with defined limits on exceptions. - ECB-level exceptions (paragraph 4) permit use for Community statistics (per agreement with the European Commission) or by scientific research organizations if individual identification is rendered impossible and the national authority that provided the information gives explicit consent. - NCB-level exceptions (paragraph 5) permit use for national or Community statistics (per agreement with national statistical authorities and NCBs), prudential supervision, “use for purposes other than those specified in the statutes,” and forwarding to scientific research organizations, all subject to non-identifiability and explicit BdF consent. - BdF staff confidentiality: - BdF staff observe professional secrecy pursuant to Article 14 of its Statute (Article L142.9 of the CMF) and Articles 226-13 and 226-14 of the Criminal Code. - Authorized recipients of professional-secrecy-covered information are defined by Article L631.1 of the CMF. - Access to individual data is limited to staff responsible for checking such data; temporary access may be granted only to named, accredited employees conducting a study of individual data. ### 0.1.4 Legal mandate and measures to encourage response - Legal framework and enforcement: - The 1984 Banking Law (Article 40) empowers the CB to determine the list, format, and deadlines for documents and information to be submitted to it. - BdF and CB closely monitor submission deadlines and maintain ongoing contacts with respondents; lateness penalties may be applied by the CB. - BdF authority to collect statistics derives from Article L144-1 of the CMF, Article 4.4 of ECB Regulation No. 2001/13, and Article 3.2 of ECB Regulation No. 2001/18 (November 22, 2001); conditions are set in Decision No. 02-01 (May 22, 2002) and Notices Nos. 02-01, 02-02, and 02-03. - Statistical collection on behalf of the CB is carried out under Article L613-8 of the CMF and regulations of the Banking and Financial Regulation Committee. - Sanctions and compliance mechanisms: - Article 34.3 of the ECB Statute vests the ECB with the power to sanction failure to observe reporting requirements; EC Regulation No. 2532/98 clarifies ECB powers regarding sanctions. - Council Regulation No. 2532/98 provides that the ECB may impose fines up to €500,000 and daily lateness penalties of €10,000. - Article 7 of Regulation No. 2533/98 specifies sanctions for reporting infractions by reporting staff: - daily fine of €10,000 (maximum €100,000) if no statistical information is forwarded by the deadline, - fine of up to €200,000 if statistical information is incorrect, incomplete, or not in compliant format. - Article 6 provides for a verification duty and mandatory collection; ECB and NCBs have right of physical access to premises of delinquent respondents; obstruction may be sanctioned by a fine up to €200,000. - In serious breaches, reporting officers may be partially or completely barred from monetary policy operations; a protocol on practical application of such sanctions is being developed within the ESCB. - Burden minimization and cost/benefit: - Regulation (EC) No. 2533/98 (Article 3 and twenty-third Whereas clause) requires the ECB to contain compliance costs by using existing statistics where possible and to consider exemptions for certain respondent categories, and calls for cooperation with the Commission to reduce reporting burdens. - BdF states that the definition of data for euro-area monetary statistics underwent a prior formal cost/benefit analysis, asking the ECB Council to rank datasets by importance and NCBs to classify datasets by cost on a scale of 1 to 7; ECB aggregated results and engaged in arbitrage. - SASM conducted negotiations with resident credit institutions during this process; credit institutions largely cooperated and source-data provision has been characterized as excellent. *Source: _cr03339 - 0.1 Legal and institutional environment* ### 0.2 Resources ### 0.2 Resources ### 0.2.1 Staff, financial, and computing resources are commensurate with statistical programs of the agency - At the BdF, arbitrage is carried out annually for the allocation of human and financial resources during the budget preparation phase, on the basis of an updated five-year projection within the framework of a rolling business plan. - A forecast of staffing levels by function and by level of responsibility makes it possible to maintain staffing at the level required to perform the tasks in question. - Staffing levels at both the SGCB and the SASM are adequate in terms of the workload, from the collection of source data through the preparation of the overall statement for resident MFIs. - The SASM reports that it has observed that the priority accorded to requests for resources for source data collection for the monetary statistics had considerably increased, inasmuch as the requests from the ECB to the NCBs are accompanied by a performance obligation and are not subject to any waiver in the view of the BdF Management Committee. - As of March 15, 2003, the SASM employs 34 regular staff and seven external technical assistants. - Of these personnel, six process data for the monetary balance sheets of the credit institutions. - Five draw up the statistics on the UCITS. - Five collect data on interest rates and make use of the survey on the distribution of bank credit. - Gathering data from respondents is completely computerized, while the collection of data on credit institutions (BAFI) is carried out initially by an SGCB unit. - SASM staffing was recently increased by five employees in order to deal with the expanded workload associated with enhanced collection of source data on bank interest rates and by one employee for the preparation of the consolidated statement on resident credit institutions. - The SASM head reports that the current staffing level is satisfactory for performing the unit’s current work plan. - In-depth studies conducted on the basis of the figures produced are expected to be developed over time; at present, they are prepared by two junior economists, supervised by a senior economist as well as, when necessary, by selected staff responsible for data production. - The SASM is currently staffed by seven economists with advanced university degrees, five supervisory staff with university degrees, and 20 statistician assistants most of whom have diplomas beyond the baccalaureate level. - The qualifications of staff are well matched with the missions entrusted to them. The backgrounds required at recruitment include solid statistical, computer, and bank accounting knowledge. - Supplementary training is provided, aimed at updating their theoretical and practical knowledge by means of training programs delivered by the BdF Training Institute or outside organizations. - The supervisory personnel regularly provide information on the methodology and procedures for preparing monetary statistics, including analysis of the legal provisions governing the preparation of such data. - Specialized ad hoc reinforcements are required from time to time, in particular for the development and maintenance of computer software. - The adaptation of computer resources to changing requirements and to improving the performance and security of procedures is ongoing, illustrated by several IT projects under way in late 2002, under the coordination of steering committees, and by their inclusion in a master plan for IT development at the BdF. - Computer hardware is regularly adapted to changing needs to ensure efficient processing and management of data. - Each employee has an individual, high performance work station. - Additional work stations are devoted to specific logistical and communications activities. - The central database is replicated in part on the local network within the DESM to permit ongoing access to the data in the event the central site is unavailable. - The software used is of recent vintage and is regularly maintained so that it offers all required functionalities. - Premises are sufficiently spacious and suitably furnished. A specially tailored technical space houses all the work stations intended for the receipt and dispatch of large files and for maintaining secured connections with other sites inside and outside the BdF. - At the BdF, budgets are prepared using ascending procedures. Capital expenditure and sizable software development efforts are financed and carried out in the context of projects, in some cases spanning more than one year. - Once the strategic arbitrage has been carried out, the financing of the multiyear projects included in the development plan is provided in full before they are launched by including funding provisions in the BdF budget. - The SASM reports that the financial resources allocated for preparing monetary statistics are satisfactory. These resources are regularly adjusted in light of changes in the workload and were significantly increased owing to the recent workload increase. - In-depth reviews are conducted to ensure the long-term evolution of the IT architecture for the collection and processing of monetary statistics. - The central idea is to move the computer processing in the separate units toward the use of SAS statistical and econometric software to introduce greater flexibility in complex calculations, in particular those associated with calculating bank interest rates which depend upon a changing sample of respondents. ### 0.2.2 Measures to ensure efficient use of resources are implemented - At the BdF, a five-year business plan is updated annually for each activity. This plan defines objectives which must be matched with resource proposals; arbitrage is carried out in the Management Committee of the BdF chaired by the Governor. - A system of incentives based on enterprise performance heightens awareness of the need to control costs while ensuring the best possible quality of service, as the performance of the work units is expected to improve each year. - Work procedures and technical resources are streamlined on an ongoing basis in view of the need to cover gradually increasing costs and obligations with material resources that are rising at a slower pace and steadily declining staffing levels. - Recourse is increasingly made to available techniques for achieving productivity gains. - Contacts with the other NCBs and with other national or foreign organizations that produce statistics make it possible to assess the BdF’s performance in producing the data necessary for the monetary statistics of the euro area. - Periodically, the allocation of tasks among the various BdF units is reassessed to avoid duplication of work and to promote the sharing of responsibility for achieving the greatest possible overall efficiency. - The unit responsible for managing the Monetary and Economic Series Database (BSME) was transferred from the Directorate of Economic Documentation and Publications to the DESM in March 2001. - The BdF is subject to annual inspections coordinated by the Audit Office, which conducts investigations of specific sectors of activity with a view to ensuring the proper use of public funds. The “monetary statistics” activity has been reviewed several times by the Audit Office. - SASM supervisors, in coordination with the DESM, promote a strategic vision and orientation embraced by staff. - Unit meetings are organized on a regular basis to inform staff of developments with ongoing activities and projects. - Meetings are devoted to the preparation of the Business plan and to the definition of local objectives with a view to awarding an annual incentive bonus, the amount of which depends upon the achievement of qualitative or quantitative performance targets identified in advance. - Compliance with standards established by BdF Management and by the Information Technology Directorate for the conduct of projects and the introduction of standardized, secure software guarantee the quality and security of processes implemented. - Efficiency gains are sought through process automation (for example, the automatic printing of letters providing notification violations of the regulations on the submission of statistical statements to the ECB). - Data preparation includes multiple levels of processing to check the accuracy of the data: internal consistency of figures, cross-document checks, auditing of revisions, and checking of certain statistical analysis documents to eliminate outliers. - The BdF regularly, on its own account or on behalf of the ECB, conducts audits to ensure the quality organization of its units and the suitability of resources for the missions assigned to them. At the request of the ECB, audits of the “monetary statistics” activity were conducted in 1999 and 2002. - Work methods are periodically reviewed to ensure improvement (for example, the wider use of SAS software). - Staff performance is appraised annually to prepare decisions on promotions and salary increases. Staff benefit from individual professional training courses. - Budgetary procedures have been adapted to allow for the use of external technical assistance and to fund expenditures outside the current budget year to address requirements for implementing new technical and computer resources in response to changes in data collection modalities. - The SASM is introducing greater plasticity in the data processing and dissemination processes: - Introduction of a laboratory for testing applications before integration into production processing. - More intensive use of SAS software (capacity to perform calculations on a large number of time series, variance calculations, detection of outliers, calculations of forecasts, and smoothing). ### 0.3 Quality awareness #### 0.3.1 Processes are in place to focus on quality - SASM supervisors recognize data quality as a complex concept involving multiple criteria and devote great attention to data quality in the broad sense. - The BdF Inspectorate General has a unit responsible for auditing central units. This unit draws up reports, including recommendations, on the operation of headquarters units, which it visits periodically. - There is a highly formal procedure for tracking implementation of the recommendations; any recommendation not implemented by the audited unit must be duly justified. - The SASM was audited in 2002; the report recommended preparing to receive computerized data from the postal system owing to the recent introduction of a structure for managing the funds gathered by the CCPs (Efiposte). - Multiple interrelations exist between the statistics producing units and between them and the study and research units; statistics from a given sector are regularly compared with those from other sectors to identify inconsistencies rapidly. - The ECB also has statistics-using or -producing units for all macroeconomic sectors, enabling verification of overall consistency for the euro area. - Within the DESM, a supervisor is designated to conduct quality assessments and centralize thinking on quality improvement. - He maintains an updated self-appraisal document used to define “quality” actions. - With his counterpart from the balance of payments area, he co-chairs a “quality” network which has submitted a report to BdF management. - The BdF Management Committee requested this report to buttress the budget requests made by the SASM before announcing decisions on their arbitrage. - The report contains observations on the risks of defective monetary statistics to the BdF’s reputation. #### 0.3.2 Processes are in place to monitor the quality of the collection, processing, and dissemination of statistics - At the SASM, successive systematic controls are applied at various stages of production: - Immediate reminders in the event of delays and reporting of infractions to the ECB. - Automatic checks upon receipt of statements and rejection if the accounting is not in compliance with standards. - Consistency tests based on changes recorded in the past and, if necessary, on the statistical distribution of individual data to detect outliers. - When anomalies are found, respondents are contacted immediately for explanations by SASM staff or by the SGCB unit responsible for collecting data from credit institutions; requests for justifications are sent by letter when necessary. - Aggregated data are checked and analyzed at various hierarchical levels. - The ECB performs checks on the accuracy of data submitted by NCBs and returns to them to seek explanations in the event of anomalies. - As part of the local security policy implemented at BdF level, a table listing and evaluating risks associated with possible problems at various stages of data collection, processing, and dissemination is prepared and regularly updated. - The method adopted, AMARIS (Approche pour la maîtrise des risques—Approach for risk control), involves six stages: description of processes; identification and evaluation of the inherent risks; identification and evaluation of the internal control mechanism; evaluation of residual risks; reports and definition of action plan; periodic activation of the five preceding stages for updating purposes. - Within the DESM, the immediate use of data for internal reports, publications, and studies helps focus on relevance and accessibility criteria and contributes to data quality. - Within the Eurosystem, regular discussion fora consider ways of improving the quality of statistics and give rise to changes in procedures when warranted. - Example: the ECB organizes an annual working meeting on methods for seasonally adjusting time series, at which the NCBs exchange experiences and make proposals. - The ECB sends the NCBs mandatory directives aimed at increasing or maintaining the best possible data quality level and developing the most relevant production strategies. - In France, the BdF receives ample feedback from national users through working parties, colloquiums, seminars, and bilateral contacts with academics and banks’ research units. - The CNIS “Money, finance, and balance of payments” group, chaired by an independent university professor and with a DESM economist serving as rapporteur, enables the SASM to keep abreast of national user community desiderata. #### 0.3.3 Processes are in place to deal with quality considerations, including tradeoffs within quality, and to guide planning for existing and emerging needs - The BdF and the ECB recognize tradeoffs between various dimensions of data quality. This arbitrage is carried out in ECB committees in which the BdF participates fully. - Tradeoffs between the accuracy of statistics and the time lag for obtaining them are examined in advance of any introduction of new collection or change to existing collection, initially within the Eurosystem and subsequently within the BdF. - Once arbitrage has been carried out, procedures are strictly defined and delineated by the regulations to achieve the objective sought. - Improvement of data quality is an important factor in criteria for drawing up the medium- and long-term work programs of the ECB and the BdF (BdF Business plan). - Users of statistics are informed of steps being taken, in particular within the CNIS “Money, finance, and balance of payments” group, where user community representatives can express their views. - Meetings are organized regularly within the Eurosystem with decision makers and other data users to analyze monetary statistics and identify new data requirements. - Numerous contact points are indicated in publications and on websites for users to express desiderata. - Relevant observations on monetary statistics could be made in working parties on macroeconomic subjects, in particular those organized by the National Credit and Securities Council (CNCT) until its suppression in July 2003. - This possibility will be maintained, in the field of household monetary and financial holdings and financing, by the new Financial Sector Consultative Committee (CCSF), stemming from the merger between the consultative committees of the CNCT and the National Insurance Council, which will be in charge of consumers matters concerning banking, insurance and securities. *Source: _cr03339 - 0.2 Resources* ### 1.1 Professionalism ### 1.1 Professionalism ### 1.1.1 Statistics are compiled on an impartial basis - The ECB and ESCB Statutes, like that of the BdF, establish the independence of these institutions from the public authorities, guaranteeing absence of interference from other public agencies in respect of the content or publication of statistical information. - Article 7 of Protocol No. 18 on the Statute of the European System of Central Banks and on the ECB: “In accordance with Article 108 of this Treaty, when exercising the powers and carrying out the tasks and duties conferred upon them by this Treaty and this Statute, neither the ECB, nor a national central bank, nor any member of their decision making bodies shall seek or take instructions from Community institutions or bodies, from any government of a Member State or from any other body. The Community institutions and bodies and the governments of the Member States undertake to respect this principle and not to seek to influence the members of the decision making bodies of the ECB or of the national central banks in the performance of their tasks.” - Article 3 of the ECB Code of Conduct governs relations between its staff and the outside world, reiterating and expanding the principle of independence. - Appointment terms: - ECB Board members, including the President and Vice President, are appointed for nonrenewable eight-year terms. - BdF: Governor and two Deputy Governors are appointed for six-year terms that are renewable once; the six other members of the Monetary Policy Council are appointed for nonrenewable nine-year terms. - Recruitment and human resources: - Professional qualifications play a decisive role in recruitment policies of the BdF; a minimum level of education is required for each qualification level and function; for the most part, employees are recruited by competitive examination. - Available positions in statistical units are announced to all staff; selection based on candidates’ intellectual and professional qualities, degree of motivation and competence, and experience. - Promotions are based on an annual appraisal; bipartite commissions draw up promotion lists applying criteria that increasingly weight selection-related factors over seniority. - Training and capacity building: - A continuing training program is conducted via the BdF Training Institute or outside organizations. - In 2002, SASM employees participated in 30 professional training courses for a total of 183 days and in orientation sessions and programs for a total of 44 days. - Outreach and methodological transparency: - The BdF posts regularly updated monetary statistics methodology on its website. - The BdF organizes numerous colloquiums and seminars targeted at universities, other central banks (in particular French-speaking countries) and various economic and professional sectors. - Through its International Banking and Finance Institute, the BdF is organizing a colloquium entitled “European Experience with Statistical Harmonization,” to be held from July 1 to 4, 2003. - The BdF has established a Monetary and Financial Research Foundation that awards fellowships to researchers and monitors their progress. - Quality control: - Research and analyses performed by the SASM are evaluated by the BdF hierarchy and analyzed by a review committee for publications (Bulletin mensuel, Revue de stabilité financière). ### 1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations - Data harmonized at the euro area level are defined in detail by ECB regulations, by guidelines defining methods for the NCBs, and by guidance notes providing further clarification. - Methods selected are determined by committees and working parties comprising experienced economists and statisticians. - Reasons behind methodological choices are indicated in the above documents; details are provided to the public in introductory notices accompanying monetary statistics tables. ### 1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics - The BdF disseminates an annual report on its website entitled “Les évolutions monétaires en 200...” (Monetary developments in 200...) that includes summary explanations and a detailed methodological notice. - The BdF monitors press articles on monetary and financial issues through its Documentation Center. - Under a service contract, the Communications Directorate receives a daily press summary of all articles relating to the BdF, the ECB, and their activities. - In cases of erroneous commentary or information, the pertinent journalist is contacted to provide factual information or may be put in direct contact with the unit concerned; in the most serious cases, the Communications Directorate may exercise the right of response set forth in the law on the press. - The Communications Directorate emphasizes maintaining close and ongoing contacts with the press to prevent incidents. ### 1.2 Transparency #### 1.2.1 Terms and conditions under which statistics are collected, processed, and disseminated are available to the public - The BdF publishes all decisions regarding the collection of monetary statistics in its Bulletin officiel mensuel. - The Recueil des textes relatifs à l’exercice des activités bancaires et financières, prepared by the BdF and the CB and published annually, presents legal provisions on monetary statistics, including Banking Law No. 84-46 and associated confidentiality rules, CB regulations, decisions of the BdF Governor, and notices from the Director-General of Research and International Relations. - Legal provisions governing ECB powers are published in the Official Journal of the European Communities; ECB regulations appear in the Monthly Bulletin of the ECB; these are assembled in a compendium updated annually and accessible on the ECB website. - All regulations on statistical issues for the euro area or relating to France are available on the BdF website. - The BdF informs the CNIS of any changes in statistical collection or new surveys introduced; the CNIS publishes this information in its annual report. - The qualitative survey on the distribution of bank credit conducted by the SASM with 15 resident credit institutions was submitted to the CNIS. #### 1.2.2 Internal governmental access to statistics prior to their release is publicly identified - The BdF indicates that government employees do not have access to the data prior to their release. #### 1.2.3 Products of statistical agencies/units are clearly identified as such - The origin of released data is clearly identified; the source (BdF, ECB, etc.) is indicated beneath each table on the website or in printed publications. - For tables with figures from various sources, relevant sources are indicated at the bottom of the table. - When figures are processed by the BdF but not collected by it, the data source is indicated followed by the statement "processed by: BdF." - The BdF’s Bulletin mensuel includes a statement on the back of the cover page restricting reproduction without express authorization of the Banque de France, referencing Article L. 122-5. 2 and 3 a) and Article L.122-10 of the Intellectual Property Code. - The BdF website home page contains hyperlinked statements under “Information from the publisher:” prohibiting electronic reproduction, electronic redistribution and creating archives without prior permission. #### 1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques - Users are informed in advance of major changes or new data collection either by ECB press releases or by articles in the ECB’s Bulletin mensuel. - The BdF website includes a heading “Travaux préparatoires” (preparatory work) to inform and consult respondents about ongoing reforms. - Disseminators of data are notified in advance of such changes; the BdF periodically informs the CNIS of substantial changes and every four years sends its medium-term statistical work program. - Increasing amounts of methodological information are posted on the BdF website for public access. ### 1.3 Ethical standards #### 1.3.1 Guidelines for staff behavior are in place and are well known to the staff - Directives defining ethical rules: - For the Eurosystem: Council Regulation (CE) No. 2533/98 of November 23, 1998 concerning collection of statistical information by the ECB. - For the BdF: Article L142-9 of the CMF, indicating staff may not communicate professional information to third parties; failure punishable under Articles 226-13 and 226-14 of the Criminal Code. - Obligations are stipulated in Guideline ECB/1998/NP28 of December 22, 1998. - In late 2002, the BdF adopted a financial code of conduct applicable to all employees; published in the BdF’s Bulletin, approved by the General Counsel of the BdF and by the MINEFI, and has the force of law. - An ethics officer was named in 2002 to oversee implementation of the code and to advise staff on conflict-of-interest risks; every BdF employee received a personal copy of the financial code of conduct. - The principle of independence in the Statutes protects staff from political interference. - Public confidence: - Opinion polls carried out at the behest of the BdF show that about 70 percent of the French population states that it is confident that the BdF will properly carry out its missions. - Personnel regulations and the BdF Statute include ethical principles (notably professional secrecy); failures may be severely punished as provided in personnel regulations and the Criminal Code. - SASM staff and incoming personnel must take cognizance of a document on local security policy (updated annually) and the ECB document on confidentiality rules pertaining to individual statistics; reminders of these rules are periodically disseminated in individual e-mails. ### 2. Methodological soundness ### 2.1 Concepts and definitions #### 2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices - The analytical framework for monetary statistics is broadly in line with the recommendations of the Monetary and Financial Statistics Manual (MFSM). - The SASM draws up: - An analytical statement of the BdF as the French component in the analytical statement of the Eurosystem, a consolidated statement of all monetary financial institutions (credit institutions that create money, money market funds, demand deposits with the postal system, and general government deposits other than central government deposits with the Treasury) by applying a residency criterion expanded to the euro area as the French component of the consolidated statement of all MFIs of the zone. - National statements (central bank and MFIs other than the BdF) by applying French territory as the residence criterion; no consolidation is performed between these two statements. Both statements are published in the BdF Bulletin in the Statistics section. - The SASM also calculates the total domestic debt of resident nonfinancial agents. - M3 definition for the euro area: - Consists of means of payment (currency in circulation and demand deposits), deposits that are reimbursable with advance notice of three months or less, time deposits at an initial term of two years or less, repos, securities of money market funds, money market instruments, and debt instruments at an initial term of two years or less. - Only positions of euro area residents that are neither MFIs nor central government institutions are included in M3. - All assets included in the composition of the M3 aggregate are regarded as monetary. ### 2.2 Scope #### 2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices - The BdF collects monthly statements from all credit institutions, that is, from all resident financial companies and quasi-companies all or part of whose liabilities are included in M3 broad money, as well as all money market funds in accordance with the MFSM. - At end-February 2003, France had over 1,000 credit institutions contracting monetary liabilities and 600 money market funds. - The BdF collects additional data quarterly on figures from OFIs, or other financial institutions, excluding insurance companies and pension funds. - Institutional coverage includes head offices of institutions which gather deposits and all resident branches, as well as resident branches of foreign banks. *Source: _cr03339 - 1.1 Professionalism (PDF chapter/section).* ### 2.3 Classification/sectorization ### _cr03339 - 2.3 Classification/sectorization ### Classification and sectorization framework - Institutional units are grouped into sectors in accordance with the requirements of the MFSM. - Sectorization distinguishes resident agents (euro area for ECB-centralized data; French territory for national data uses) and nonresidents, and delimits sectors and subsectors. - Resident sectors defined under ECB Regulation 1998/16 (drawn up in accordance with ESA 95 and MFSM) include: - Monetary financial institutions: ECB and NCBs, credit institutions that create money, money market funds, other monetary financial institutions; - General government (central government, social security agencies, local governments); - Private nonfinancial agents: nonfinancial corporations, households, nonprofit institutions serving households; - Other financial institutions: insurance companies and pension funds; nonmonetary UCITS (mutual funds other than money market funds), investment firms, special purpose vehicles (fonds communs de créances), etc. - Noted inconsistencies in Banque de France (BdF) consolidated MFI statements: - General government is not subsectorized on the assets side in BdF balance sheet (Tableau 11 / Table 7 of the Digest) and in MFIs other than BdF (Tableau 12 / Table 8 of the Digest); consequence: operations with the central government are not identifiable, contrary to MFSM provisions. - In Table 12, operations of MFIs other than BdF with all MFIs (including the BdF) are not subsectorized; however, operations of other MFIs with the central bank are identifiable in Table 11 (Table 7 of the Digest), consistent with MFSM. - MFSM recommendations and ECB reporting requirements: - MFSM recommends collection of monthly data on claims on, and liabilities to, central government to facilitate timely high-frequency monitoring of net claims on government. - ECB requires monthly reporting of liabilities to central government but only quarterly reporting of claims on government (assets). - Balance sheet composition (assets side for MFIs): liquidity, loans, securities other than shares, securities of money market funds, shares and other equity holdings, fixed assets, and other asset items. - Liabilities side for MFIs: banknotes and coins in circulation, deposits, the securities of money market funds, debt securities issued, capital and reserves, and other liabilities. - Deposits broken down into demand deposits, time deposits, and deposits with advance notice requirements; for monetary aggregates, time deposits further broken down by initial maturity (money aggregates for the euro area do not include financial assets with an initial term exceeding 2 years). - Treatment of repos and derivatives: - Repos treated as guaranteed loans if contracts stipulate a firm commitment to unwind; if only an option to unwind is specified, that treatment does not apply. Liquidity received against pledged assets is posted under repos on liabilities side. - Only credit derivatives operations recorded on MFI balance sheets are option premiums and margin calls: options purchased, initial margin calls, and margin calls against open forward positions recorded under “other liabilities.” - This classification aligns with ECB guidelines, but MFSM recommends these be kept as separate categories on the balance sheet, outside other assets/other liabilities. - Other forward financial instruments, including swaps, are recorded as off balance sheet items, whereas both MFSM and ECB recommend that swaps be recorded on the balance sheet. ### Basis for recording — valuation of stocks and flows - Stocks: - Monetary balance sheet components of MFIs are valued in accordance with Regulation ECB/2001/13 and the “Guidance Notes to Regulation ECB/2002/13.” - Balance sheet items generally must be valued at market prices (fair value). - Exceptions to market valuation (valued at nominal value, where applicable gross of provisions for asset depreciation): banknotes in circulation, deposits, capital and reserves, MFI investments in notes and loans. - Leasing operations must be valued as loans subject to amortization schedules. - ECB allows NCBs to use certain national accounting rules when they differ from market values and are applied by MFIs in their statistical reporting (examples: securities held for sale valued at acquisition cost or market value, whichever is lower; securities held for the long term and debt securities recorded with redemption values). - Loans and book valuation: - Under Eurosystem methodology, loans are valued at their book value, after deduction of provisions for anticipated losses; deposits, capital and reserves are also valued at book value. - Footnote clarification: book value is "for the remaining capital excluding accrued interest, but including provisions for depreciation." - Flows: - Valuation of flows follows ECB rules. Flows are financial transactions during the reference period, i.e., change in stocks between ends of two consecutive periods minus adjustments for reclassifications, exchange and valuation effects for securities portfolios, and debt write-off. - For money market funds, Regulation ECB/2002/8 (third Whereas clause) indicates it is not necessary to collect data on price revaluations of MMF shares/units because changes in value are considered as interest reinvested; consequently, no adjustment by the ECB is required for securities investments of MMFs. - Comparison with MFSM: - ECB regulations are very close to MFSM except MFSM calls for recording securities operations exclusively at market prices. - Practices applied by BdF and credit institutions: - BdF: balance sheet components valued at market prices except those for which ECB requires nominal value; all adjustments for flows are singled out. Same applies to money market funds (stocks and flows), except valuation effects on securities per Regulation ECB/2002/8. - Credit institutions: valuation rules required by the CB for accounting and prudential declarations. These rules depart from international statistical standards for securities operations — valuation of portfolio investment securities, investment securities, securities representative of debts and financial derivatives diverges from MFSM and ECB Guidance Notes (see paragraphs 46–48 of Guidance Notes). Nevertheless, adjustments for flow calculations are in all instances calculated in accordance with the MFSM. ### Valuation practices by instrument (as summarized in source table) - Deposits - Valuation method in monetary balance sheet of credit institutions: Remaining principal owed excluding accrued interest - Category of adjustments calculated by the Banque de France: Reclassification and exchange effects - Conformity with MFSM: Yes - Remarks: Adjustments relating to valuation effects and debt write-off not relevant in the case of deposits - Loans - Valuation method: Remaining principal owed gross of provisions excluding accrued interest - Adjustments: Reclassification, exchange effects, effects of valuation and debt write-off - Conformity with MFSM: Yes - Remarks: Monthly collection of debt write-off figures since January 1, 2003. Estimates on the basis of the half-yearly profit and loss account for the prior period - Trading securities (assets and liabilities): securities held with the intent to resell within six months - Valuation method: Market value, including accrued interest - Adjustments: Reclassification, exchange effects, valuation effects, and debt write-off - Conformity with MFSM: No for stocks; Yes for flows - Remarks: Monthly calculation of valuation effects on the basis of security-by-security collection since January 1, 2003. Estimates for the prior period - Securities held for sale and portfolio securities: securities held for six months or more - Valuation method: Acquisition cost or market value, whichever is lower - Adjustments: Reclassification, exchange effects, valuation effects, and debt write-off - Conformity with MFSM: No for stocks; Yes for flows - Remarks: Calculation of valuation effects effective with reference date March 2003 - Investment securities: securities held until their maturity or the subject of full and complete endorsement as to term or rates - Valuation method: Redemption value with amortization of the gap between the acquisition cost and the redemption value over the residual life of the security - Adjustments: Reclassification, exchange effects, valuation effects, and debt write-off - Conformity with MFSM: No for stocks; Yes for flows - Remarks: Calculation of valuation effects beginning with reference date March 2003 - Equity securities - Valuation method: Acquisition cost, gross of provisions for depreciation - Adjustments: Reclassifications, exchange effects, and debt write-off - Conformity with MFSM: No for stocks; Yes for flows - Remarks: Calculation of valuation effects pointless as provisions for depreciation are included in stocks (constant value in the monetary balance sheet) - Debts in the form of securities - Valuation method: Redemption value - Adjustments: Reclassifications, exchange effects, and debt write-off - Conformity with MFSM: No for stocks; Yes for flows - Remarks: Calculation of valuation pointless as these debts are recorded at their redemption value - Capital and reserves - Valuation method: Face value - Adjustments: Reclassifications, exchange effects, and debt write-off - Conformity with MFSM: Yes - Remarks: Recording of adjustments on debt write-off required by the ECB to bring balance sheet into balance ### Recording conventions and procedures - Currency conversion: - Conversion of transaction currency into euros carried out at the market exchange rate prevailing on the date of the transaction (ECB regulatory provisions). - Conversion of stocks of financial assets and liabilities denominated in foreign exchange uses the interbank market rate observed on the closing date of statements (Article 4 of CB Instruction No. 94.05 of March 14, 1994 on the accounting of operations in foreign exchange). - UCITS accounting: - All assets valued at market price, including derivatives and futures. - Accrual basis (2.4.2): - Interest on financial assets and liabilities is determined on an accrual basis over the accounting period and includes interest accrued but not paid. - Interest arrears are not recorded separately if initially recorded on an accrual basis before maturity; once due, interest continues to be included in the corresponding obligation. - Payment arrears on commissions or similar charges associated with off-balance-sheet financial instruments are recorded in credit institutions’ accounts receivable or payable. - Claims classified as overdue are financial assets; off-balance-sheet instruments are not regarded as financial assets. - Transactions are recorded at the time they occur (simultaneous recording) — i.e., when there is a transfer of ownership, based on market prices at which assets/liabilities are bought or sold; transactions are recorded at issue of new securities or transfer of securities. - Grossing/netting (2.4.3): - Data are collected and drawn up gross. - Provisions for bad debts are recorded under “capital and reserves.” - Claims and liabilities in respect of a nonfinancial agent or group of agents are never offset in banking institutions’ accounts, except in the case of a formal merger agreement (in which case term and currency for accounts submitted for the merger agreement must be identical). - These grossing/netting practices are consistent with MFSM requirements. *Source: _cr03339 - 2.3 Classification/sectorization (PDF chapter).* ### 3.1 Source data ### _cr03339 - 3.1 Source data ### 3.1.1 Source data collection and respondent liaison - Source data are collected from comprehensive data collection programs that take into account country-specific conditions. - Accuracy ensured through in-depth cooperation between the SASM, the SGCB, and respondents. - SGCB disseminates comprehensive documentation to respondents on completing reporting documents (BAFI database of financial agents); documentation is updated frequently. - Mandatory electronic submission; data are automatically checked on receipt and immediately rejected if anomalies are observed to allow resubmission within deadlines. - Changes in submission formats are announced in advance and generally discussed in working parties with respondents (including AFECEI). - Technical meetings are organized for IT units supporting respondents’ accounting units. - SASM maintains contacts and periodic meetings with UCITS managers; on-site surveys planned in selected institutions to verify quality (SASM requested Inspectorate General of the BdF to include such surveys in its work program for the second half of 2003). - Respondents can receive, on request and free of charge, detailed monetary statistics: about 2,000 monthly time series and 2,200 quarterly time series in electronic form. - Credit institutions report via special statements and accounting statements received in the BAFI database managed by the SGCB. - All institutions must submit quarterly statements (Series 4000). - About 350 credit institutions are required to submit monthly statements for monetary statistics purposes (Series 8000). - As provided under Article 2.2 of Regulation ECB/2001/13, the total of their combined balance sheets represents more than 95 percent of the total balance sheet for all credit institutions. - Money market funds: information collected security line by security line and by counterpart agent; monthly reports submitted directly to the SASM; monthly coverage represents more than 95 percent of the activities of the total population. - Deposits of agents other than MFIs in the central government collected per Annex IV of Guideline ECB/2002/05 using the summary statement of Treasury operations (SROT) published monthly by the MINEFI; when SROT not available, SASM uses Weekly Treasury Statement; postal system deposits collected directly. - Corrected declarations from reporting agents are systematically used after anomaly identification. - Banking sector restructuring (mergers, acquisitions, transfers) is monitored for updating the list of MFIs published by the ECB; SASM receives copies of licenses approved by CECEI (secretariat services by BdF’s DECEI). - Special statistical declaration introduced effective early 2003 for transfers of assets to funds that do not collect savings from the public (statement 8026) in implementing Regulation ECB/2001/13. - Coordination with banking sector via SGCB as single point of contact; planned periodic gatherings of reporting bank representatives to examine implementation issues of the specific regulation on collection of statistical information by the BdF. - For UCITS, meetings of reporting agents, COB, and SASM are organized twice yearly. - The Head of the SASM and deputy are members of the Working Group on Money and Banking Statistics of the ESCB. ### 3.1.2 Conformity to definitions, scope, classifications, valuation, and timing - Globally, data collected meet statistical requirements. - Reporting obligations reference a chart of accounts and customer nomenclature common to the CB and the BdF, integrating statistical definitions into credit institutions’ information systems. - Statistical breakdowns of money market funds’ securities portfolios are prepared by SASM using ISIN codes and a master list (référentiel) derived from Balance of Payments. - In monthly statements (Series 8000) from credit institutions, insignificant accounting headings are recorded under “other assets” or “other liabilities” without counterpart breakdown; by contrast, all basic headings of quarterly statements (Series 4000) and monthly statements from money market funds are treated separately regardless of significance. - Postal service information comes from nonaccounting sources; effective July 2003, banking activities of the Postal Service will be integrated into declarations from Efiposte, which will be regarded as an MFI and subject to credit institution reporting requirements. - Since the introduction of euro cash in early 2002, the “banknotes in circulation” item for each NCB is calculated by applying the percentage of ECB share capital held by each NCB to total banknotes in circulation in the euro area, after deduction of an 8 percent share allocated to the ECB. ### 3.1.3 Timeliness - BdF Governor’s Decision No. 02-01 of May 22, 2002: monthly declarations from credit institutions and money market funds must be provided not later than ten business days following the closing date. - Procedures for systematic reminders introduced; all declarations must be submitted electronically. - Plan to systematize electronic messaging with respondents. - The General Accounting Directorate of the BdF forwards information for the BdF’s statistical balance sheet between five and ten business days following the closing date. ### 3.2 Statistical techniques - Monetary statistics compiled exclusively in a computerized environment on a large system composed of collection applications (BAFI, UCITS), processing/calculation applications (SMF and, as from 2003, SISTRe), and an audit/infractions monitoring application (IDSMF). - BdF’s economic database unit (BSME) forwards series to the ECB and posts them on the BdF’s website. - Recent focus on improving statistical calculation procedures using SAS software. - All statements from credit institutions use a standardized format and standardized terminology; a common file format must be used by respondents. ### 3.2.2 Data adjustments, estimation, and seasonal procedures - Missing data estimated by: - Carrying data forward from the latest declaration for small credit institutions, or - Using ARIMA type forecasting models for large respondents. - For money market funds, the data collection application prepares estimates based on the net asset value declared daily to the COB and the last known breakdown of that net asset value. - Beginning in early 2003, SASM will seasonally adjust the series of greatest importance for analyzing the national monetary situation using methods based on research presented at an ESCB workshop. ### 3.3 Assessment and validation of source data - Electronic submissions are subject to three types of checks: - Formal checks (file structure), - Internal consistency checks (declarations), - Statistical controls (outlier detection via time series examination and contribution analysis). - In accordance with Article 5 of Decision No. 02-01, reporting agents must explain anomalies as rapidly as possible; corrected declarations must repeat all information for affected reference dates. - All anomalies resulting in contacts with respondents are logged. - For interest rate statistics on new contracts collected by sampling since January 2003, the list of reporting institutions will be reviewed once a year (Paragraph 24 of DGEI Notice No. 02-01 of May 23, 2002). - Under regulations ECB/2001/13 and ECB/2001/18, individual information on securities of credit institutions (monthly) and new lending (quarterly) will allow partial verification of respondents’ customer sectorization. ### 3.4 Assessment and validation of intermediate data and outputs - Prior to dissemination to the ECB or publication, summary tables focused on the contribution of resident MFIs to M3 are prepared on the basis of MFI balance sheet series for cyclical analysis of collection. - Comparisons made with data from Balance of Payments and financial accounts; monthly monetary statements systematically checked against prudential or statistical documents submitted quarterly or more frequently (see Paragraph 24 of DGEI Notice No. 02-01 of May 23, 2002). - Trends for claims and obligations among resident MFIs are monitored for structural bias; major headings in aggregated MFI balance sheet (loans, deposits to/from agents other than MFIs, stock of debt securities issued) are analyzed. - Origins of pronounced shifts are identified and analyzed using summary tables; total loans to enterprises declared by credit institutions are compared with amounts from the Credit Information Office (roster of all loans to enterprises of €76,000 or more). ### 3.5 Revision studies - A study is under way to identify emerging trends and patterns in monetary statistics since 1999 for seasonal adjustment procedures. - Central bank data are not revised; data for other MFIs are regarded as provisional and susceptible to subsequent modification when new information becomes available. - Revisions arise from integration of late-reporting institutions’ declarations, correction of identified anomalies, or cross-checking monthly statements against quarterly statements (Series 8000 vs Series 4000). - Revisions introduced for the end of a given quarter based on comparisons of monthly statements against exhaustive quarterly statements are used to correct declarations for subsequent months; work initiated to perform these corrections by sampling cross-check procedures. *Source: _cr03339 - 3.1 Source data* ### 4. Serviceability ### 4. Serviceability ### 4.1 Relevance - The relevance and practical utility of existing monetary statistics for monetary authorities are regularly reviewed in light of comments and requests for supplementary analyses. - ESCB procedures provide for consultation with all users before changes to statistical regulations (cost and merit exercise); this was carried out for the first time in 2001 and led to promulgation of Regulations 2001/13 and 2001/18 on the consolidated balance sheet of the MFIs and on interest rate statistics, respectively. - The SASM participates in the Working Group on Money and Banking Statistics at the ECB and can present views on methodological changes proposed by international organizations. - Twice a year, the CNIS invites research units of local banks and sends them medium-term work programs prepared by all public statistics producing units. - The SASM adopts a systematic approach to analytical, reporting, study, and research work intended for publication to heighten awareness of economists’ and decision makers’ needs. - The SASM has collaborated within the BdF with the Directorate of Econometric Studies and Research and the DESM to develop a model explaining the demand for money in the euro area. ### 4.2 Timeliness and periodicity - Timeliness: - In accordance with the SDDS, the analytical account of the central bank is published on the BdF website within 15 days following the end of the reference month. - The analytical account of the MFI sector is published on the BdF website within 30 days of the end of the reference month. - Monthly results of the breakdowns of deposits and loans by counterpart agent and category of instrument are published within 30 days. - National data and components of the money supply aggregates of the euro area are not published by the BdF until after the ECB has published the monetary statistics of the euro area. - Periodicity: - In accordance with the SDDS, the monetary balance sheet of the central bank and those of the other MFIs are prepared and disseminated monthly. - Some detailed information is published quarterly (for example, the breakdown of deposits and loans by country and currency). ### 4.3 Consistency - Internal consistency: - Statistics show a high degree of internal consistency due to multiple checks on source data. - Declarations from credit institutions and the central bank regarding reciprocal operations are consistent overall. - Accounting identities in submissions of credit institutions and UCITS are subjected to automated checking procedures when downloaded into BdF data collection applications; verification is required for use by user units. - Statistics on transactions in shares of money market funds are checked by comparing them with change in stocks and change in share/unit values. - During preparation of financial accounts, systematic comparisons are performed between monetary statistics and accounts of counterpart agents. - Systematic internal validation procedures verify consistency of published results by comparing summary statements relating to monetary and financial statistics. - Consistency over time: - Since July 1998 for credit institutions, and since March 1999 for UCITS, monetary statistics are consistent with ECB requirements and therefore homogeneous. - For the prior period beginning in March 1993, the BdF disseminates series in the current format using backward extrapolation; methods used to reconstitute these series and ensure compatibility are presented in the quarterly monetary statistics methodology on the BdF website. - Statistical series are therefore consistent from 1993 forward. - Consistency with other data sources/frameworks: - Systematic comparisons are made between declarations received by Balance of Payments from credit institutions regarding operations with nonresidents and their monetary and prudential statements. - Major discrepancies are followed up by checking with respondents. ### 4.4 Revision policy and practice - Revisions schedule and practice: - In accordance with Annex XII of the ECB Guideline of July 30, 2002 (ECB/2002/5), revisions relating to the preceding month are incorporated into publications of the current month (monthly revision). - Revisions arising from cross-checking monthly statements with quarterly statements are customarily introduced not later than the publication relating to the second month of the following quarter. - Other revisions mainly stem from final closure of the national accounts and are made as necessary. - Revision procedures are disseminated to the public in the methodology available on the BdF website; when revisions are of considerable significance, reasons are stipulated. - Identification of preliminary data: - The monetary statistics methodology indicates that data for the latest period are provisional. - However, published statistical tables do not mention that the most recent data are provisional. - Studies and analyses of revisions: - The BdF does not publish results of its revision studies. - The impact of revisions is clearly indicated whenever they are significant in scope. ### 5. Accessibility (summary) - General: Data and metadata are easily available and assistance to users is adequate. - 5.1 Data accessibility: - Presentation and interpretation: - Summary analyses of monetary developments are published monthly under the “Actualités” (recent developments) heading of the BdF website as “Stat Info,” covering money supply aggregates of the euro area, the French contribution, loans granted by MFIs resident in France, and total Domestic Indebtedness of nonfinancial agents resident in France; pages include graphics. - Stat Info pages are published on the BdF website two business days after the ECB disseminates its press release on monetary developments for the month. - Monthly monetary developments are also commented in the BdF’s Bulletin mensuel, which contains accounts of the central bank and other MFIs in balance sheet form, tables of main deposit and loan series and components of total domestic debt. - Monetary developments in the preceding year are presented in a special document under the Statistics heading of the BdF website. - Dissemination media and formats: - Monetary statistics series published in the Bulletin mensuel may be downloaded free of charge from the BdF website. - Series may also be accessed via external vendors such as DATASTREAM, WEFA, and ECOWIN. - Release schedule: - The release schedule for euro area (ECB Press Release) and national monetary data (Stat Info) are on the website; the schedule indicates, beginning in November of year n, the release dates for the monetary statistics for each month in year n+1. - Simultaneous availability: - Statistics are released simultaneously to all users; first dissemination is by posting on the ECB website, and then the BdF website. - Nonpublished subaggregates: - Nonpublished (but nonconfidential) subaggregates are provided free of charge upon request, by whatever delivery method best suits the recipient. - 5.2 Metadata accessibility: - Documentation: - Information is provided in the monthly monetary statistics methodology published on the BdF website. - Differences from internationally accepted standards, in particular from the MFSM, are not highlighted in said methodology. - Metadata pages relating to the monetary statistics are also available on the IMF’s DSBB. - Levels of detail: - Efforts are made to adapt deliveries to various categories of users (economists, banking profession, academics, etc.). - 5.3 Assistance to users: - Contact information: - The preparing unit and telephone number are indicated on each table published by the BdF. - The Internet address of the SASM is indicated on the Stat Info pages. - A euro area brochure for the general public published by the BdF in French and English introduces major economic, monetary, and financial indicators and is available on the BdF website. - Catalogues and charges: - The catalogue of publications is available on the BdF website and in pamphlet form. *Source: _cr03339 - 4. Serviceability* ### 0.1 Legal and institutional environment ### _cr03339 - 0.1 Legal and institutional environment ### Legal and institutional responsibilities - The Banque de France (BdF) responsibility for collecting and processing balance of payments statistics is defined by the law of August 4, 1993 (as amended by the laws of December 31, 1993 and May 12, 1998) reforming the Statute of the Banque de France. - Article 16 of the law, reiterated in Article L.141-7 of the Monetary and Financial Code (CMF), states: “In accordance with the terms and conditions referred to in Article 15, the Banque de France shall establish the balance of payments and assess the country's net external asset position on behalf of the State and as directed by the Minister for Economic Affairs and Finance.” - Decree 89.938 of December 29, 1989 (as amended) governed financial relations with the rest of the world; a new decree of March 7, 2003 (No. 2003-196) and its implementing order revoke the 1989 decree and modify selection criteria for reporting enterprises, allowing adoption of a revised data collection system to be introduced in 2004. - The BdF is a public entity; its statistical units are recognized by the Conseil national de l’information statistique (CNIS) as public statistics producing units, but not as ministerial statistics units (SSM). - Within the Directorate-General of Research and International Relations of the BdF, the Balance of Payments Directorate (DBDP) assumes full responsibility for compilation of the balance of payments. - Data dissemination: Article 16 of the law of 1993 stipulates that the minister responsible for economy and finance shall publish such information. Since 2001 press releases on the balance of payments have been published under the dual imprimatur of the BdF and the Ministry of Economy, Finance, and Industry (MINEFI). The Treasury Directorate drafts press releases taking DBDP changes into account; statistical tables on the MINEFI website indicate both BdF and MINEFI as the source, though MINEFI is responsible only for the conception (“réalisation”) of the statistical tables. ### Data sharing and coordination - Data sharing and coordination governed largely by Law No. 51-711 of June 7, 1951 and Decree No. 84-628 of July 17, 1984; CNIS coordinates mandatory government statistical surveys, including DBDP’s survey E84 (advances received with orders and trade credit). - Agreements in place with the Directorate-General of Customs and Indirect Duties (DGDDI) to forward results of the survey of the costs of services associated with international trade; that survey is endorsed as mandatory by CNIS and was developed with the DBDP. - Frequent meetings and cooperation with the Directorate of Tourism for nonmandatory surveys providing travel data on revenue and expenditure sides. - DBDP to enter into a data sharing protocol with INSEE to establish an enterprise registry for identifying samples for a services survey planned for 2006. - DBDP has access to monthly international trade data from DGDDI; IT protocols exist between DGDDI and DBDP; meetings organized at least once each half year; informal contacts frequent and facilitated by DBDP staff with prior DGDDI experience. - Within BdF, DBDP maintains systematic contacts with DESM enabling comparisons of monetary statistics with banking sector and balance of payments statistics; internal working groups (e.g., 2002 study on operations of credit institutions with nonresidents; new group beginning work in 2003 on general government). - Observation: further integration and increased sharing of individual or aggregated data from administrative or survey sources could improve assessment and validation; joint task force established end-2002 by DBDP and INSEE to address discrepancies between balance of payments and national accounts. ### Confidentiality and use of respondent data - Confidentiality protected by the law of 1993, reiterated in the CMF; BdF employees bound by professional secrecy (defined by Law No. 83-634). Additional guarantees: BdF’s Financial Code of Conduct; penalties under Articles 226-13 and 226-14 of the Criminal Code for unlawful communication of nonpublic information. - Regulatory texts to financial intermediaries and direct reporters specify information will be used for drawing up the balance of payments. The original version of the 1951 law (amended 13 times) is incorporated in texts to financial intermediaries to reference confidentiality; texts for direct reporters do not indicate confidentiality. - Within DBDP, data access is organized on a need-to-know basis; information system access is password protected; employee accreditation defined by hierarchy per Article 7 of the Bank’s Ethics Code. - DBDP prepares a “local security policy” and a regularly updated memorandum on “the local security policy on computer and office automation applications” outlining rules, controls, and security procedures to ensure confidentiality and integrity of individual data. - Data dissemination follows CNIS notice dated June 6, 1980: for a table to comply with confidentiality rules, (i) minimum number of statistical units per cell must be not fewer than three and (ii) no statistical unit may account for more than 85 percent of the aggregate in that cell. - Annual review: Council of Governors of the European Central Bank reviews a report on implementation of the Guideline of December 22, 1998 on confidentiality of individual statistical data gathered by ECB staff assisted by national central banks; latest report referenced SEG/Gov/X/320a submitted December 13, 2002. ### Statistical reporting mandates and respondent burden management - Reporting obligations defined by the decree of March 7, 2003 and the 1951 law; decree specifies reporting obligations for residents carrying out international financial and nonfinancial transactions and BdF’s power to collect IIP data; an order specifies transaction/holding thresholds beyond which reporting applies. - Decree drawn pursuant to powers granted by Law No. 66-1008 of December 28, 1966 as amended; Article 9 of the 2003 decree introduces a new penalties framework based on Article L.165-1 of the CMF facilitating recourse to sanctions. - The 1951 law enables BdF to conduct statistical surveys after CNIS approval; mandatory surveys are subject to a penalty system managed by the Disputes Committee of the CNIS; law enables BdF to obtain results of surveys conducted by other institutions in the French statistical service once CNIS recognizes the need. - BdF considered respondent burden: DBDP established a moratorium from 1998 to 2002 on additional statistical requests for banks due to euro introduction, Y2K switchover, and introduction of euro cash. - Inspectory activity resumed in 2003: BdF Inspectorate General, assisted by DBDP experts, to resume inspection missions in major banks as in 1992 and 1995 to examine processing chains after introduction of the reporting threshold of €12,500 and reliability of portfolio investment declarations. - BdF policy to announce changes in its data collection system 6 to 12 months in advance (or 18 months for most burdensome changes); principal use of CNIS as information channel. CNIS specialized group on “Money, Finance, and the Balance of Payments” reviews the three-year statistical program at least twice a year. - CNIS emphasizes minimizing statistical burden and avoiding duplication; all changes in data collection system have been presented to CNIS since its establishment. ### Staff, financial, and computing resources - Staffing devoted to producing and disseminating balance of payments, IIP, and research: 40 economists and managers and 120 statisticians and technical and administrative support personnel. - Average length of service: 10 years for staff, and 6-8 years for supervisory personnel. - SAFIR computer application: initial portions December 1998 (master lists, some statistical processing, acquisition of validated data); expanded December 1999 to include collection and checking of data submitted by credit institutions, enterprises, etc.; since then, two to three new versions put into production each year. - Technical architecture: client/server on dedicated UNIX server; ORACLE used for master lists and source data; SAS used for statistical processing. Every staff member has a dedicated workstation and access to personal and/or shared computer resources. - Continuous development of SAFIR is a priority in view of major improvements required when collection system is revamped between 2004 and 2006. - Financial resources: cost of compiling France’s BOP and IIP reimbursed by government through the Government-Banque de France Commission established by the 1993 agreement. Three-year plan exists for staffing, budgets, and IT resources for statistical program. European work program horizons: STC prepared for one year; CMFB prepared for two years. ### Measures for efficient use of resources - Business plan, including statistics portion, widely disseminated to staff; all supervisory staff participate in preparing the statistical plan for the coming year; detailed objectives published within DBDP. - Information notes and reports of STC and CMFB disseminated; close collaboration with DESM units (SASM, SEVAM, SESOF) to enhance procedural efficiency. - DBDP regularly reviews work procedures due to frequent changes from European decisions and multiple audits; aims to identify unnecessary tasks and nonrelevant controls; control thresholds are raised as necessary. - Annual review of budgetary procedures to reanalyze and justify new requirements. ### Quality awareness and quality monitoring - A report on the quality of statistical data (outcome of network introduced February 2002 between DBDP and DESM) was submitted in October to the Banque de France Management Committee; summary published in February 2003 Bulletin de la Banque de France. The report uses the IMF (DQAF) approach adapted by European institutions and assesses progress and efforts needed on data quality. - DBDP participates in joint ECB-Eurostat task force on balance of payments quality; final report to CMFB in June 2003. DBDP volunteers in a pilot exercise with four other institutions to test quantitative indicators (series stability, internal and external consistency, etc.); some indicators already calculated periodically at DBDP. - DBDP subject to four series of audits/assessments: general internal audit of the Bank (with regular monitoring of recommendations), Eurosystem audit (methodology, procedures, IT security), Audit Office, and explanations to the Euvrard Commission (examination of compensation conditions for services rendered to government); latter two focus on costs and economies of scale. - Quality monitoring processes: DBDP uses periodically updated quantitative indicators; range to be supplemented when European-level indicators finalized. The DPDB-DESM network report and internal/external audits contribute to problem identification. - Two bodies express views on quality: ECB (and Eurostat jointly) through working groups/task forces guiding data improvement strategies; CNIS certifies rigor of survey methods and gathers user views regularly. - DBDP organizes regular meetings with major public sector users (INSEE, Treasury); ad hoc actions include colloquium in March 2002 on "French Direct Investment in Globalization: Assessment and Stakes" to take stock of concepts, figures, and statistical system evolution with wide stakeholder participation. - Tradeoffs: DBDP-DESM report focuses on accuracy/reliability versus timeliness. As part of the euro area, monthly periodicity and rapid publication are justified for monetary policy users. Quality improvements are considered in national statistical programs presented to CNIS and in European forums. Semi-annual CNIS specialized group meetings help identify new statistical requirements and gather user views. *Source: _cr03339 - 0.1 Legal and institutional environment* ### 1. Integrity ### 1. Integrity ### 1.1 Professionalism — impartiality, methods, and commentary - Legal and institutional independence - The ECB and ESCB Statutes, like that of the BdF, establish independence from public authorities and prohibit seeking or taking instructions (Article 7 of Protocol No. 18 on the Statute of the European System of Central Banks and of the ECB; reference to Article 108 of the Treaty). - Staff training and composition - DBDP implemented, in early 2003, a training and professional development course on statistical analysis for statistical directorate staff. - Staff are encouraged to participate in IMF seminars. - Preference in recruitment (competitive examinations) is for statisticians, international economists, or graduates of business schools. - Increased presence of INSEE staff on secondment to the DBDP, particularly with experience in goods, direct investment, FATS, and statistical surveys, is encouraged. - Internal review and external dissemination of research - DBDP units prepare research and studies discussed internally (including internal seminars) and with directorate-general and other Bank directorates as applicable. - Staff are encouraged to present results in European and international working parties and external colloquiums (example: international investment position study presented at the Irving Fisher Committee colloquium organized by the BIS and at the IMF Balance of Payments Statistics Committee). - Source and technique choices driven by statistics - DBDP uses many statistical models, including forecasting of Key Items at M+35; estimation for amounts below the €12,500 threshold not reported by banks; estimation of wages of cross-border workers; calculation of accrued interest; calculation of reinvested earnings; estimation of stock of direct investment; trade credit survey; and two statistical surveys being tested to estimate tourism revenue and expenditure. - Example of source-driven change: introduction of Intrastat in 1993 removed a financial framework used by an econometric model, prompting a survey on the stock of advances on orders and trade credits; survey characteristics were presented in late 1993 to the CNIS Validation Committee and received approval. - Conclusions of European task forces inform statistical techniques. - Responding to misuse or erroneous interpretation - Two internal press reviews and a press clipping service by the DBDP documentation unit and management identify interpretation errors. - Directorate-General and, where appropriate, the Governor are alerted by flash notices. - Bank communications policy emphasizes pedagogical medium-term efforts rather than individual reactions; de facto policy is not to inform the public via media of commentary on erroneous interpretations. ### 1.2 Transparency — access, identification, and change management - Public availability of terms and conditions - The 1993 law on the Statute of the Banque de France and the annual report on the balance of payments and France’s international investment position are disseminated on the BdF website, together with regulatory texts for respondents containing applicable legal provisions. - CNIS and INSEE websites may be consulted for general provisions governing French statistics. - BdF statistical publications (hard copy) indicate the website address; telephone numbers and postal addresses are beneath tables and charts (hard copy and Internet) for additional information. - Internal governmental access prior to release - BdF website information indicates BdF compiles balance of payments on behalf of the government. - Time lag between sending figures to the MINEFI and MINEFI publication of the press release is generally short but may vary (e.g., due to weekends). - Draft MINEFI press release is reviewed rapidly by the DBDP. - Only the metadata disseminated on the IMF’s DSBB clearly indicate which government employees have access prior to release and at what stage. - Identification of statistical products - MINEFI press release is published under the imprimatur of the MINEFI and the BdF; title references that data are compiled by the BdF. - BdF tables reference the DBDP as data compiler and BdF requests origin be specified when statistics are used or reproduced. - Advance notice of methodological changes - Major methodological changes are announced to user representatives in the CNIS, which disseminates this information (including on its website). - If respondents have concerns, DBDP practice is to alert them 6 to 18 months in advance to enable necessary steps. - Future methodological changes through 2006 are described in an information notice available on the Internet. ### 1.3 Ethical standards — codes and oversight - Financial code of conduct and ethics officer - BdF has adopted a financial code of conduct applicable to all employees; published in the BdF’s Bulletin officiel and approved by the General Counsel of the BdF and by the Ministry of Economy, Finance, and Industry, with the force of law. - An ethics officer was named in 2002 to surveil implementation of the code and advise staff requesting opinions. - Every BdF employee has received a personal copy of the financial code of conduct. - Statute, personnel regulations, and enforcement - BdF Statute and personnel regulations include ethical principles (notably professional secrecy); employees are informed upon assuming positions. - Failures to observe principles may be severely punished as provided in personnel regulations and the Criminal Code. - Local security policy and reminders - DBDP employees and new staff must take cognizance of the local security policy document, updated annually. - Reminders of rules are periodically disseminated by individual e-mails. --- ### 2. Methodological soundness ### 2.1 Concepts and definitions - Conformity with BPM5 and deviations - Concepts and definitions used generally conform to the fifth edition of the Balance of Payments Manual (BPM5). - Reinvested earnings of direct investment firms are reported as income and to the financial account in accordance with BPM5 recommendations; however, calculations include capital gains and losses as well as profits and losses from exceptional operations—consistent with the French accounting framework and comparable to practices in other countries (IMF study, 1997), though departing from BPM5. - Direct investment operations are recorded with a clear distinction between share capital, reinvested earnings, and other operations; breakdowns by institutional sector and industry are available for resident enterprises. - Under the direction principle, long-term loans from a direct investment enterprise to its parent are recorded as disinvestment under direct investment, per BPM5 recommendations. - Other reverse investment transactions between affiliates (share capital transactions, short-term loans, trade credit, cash flow operations) are not recorded under direct investment but under other functional categories in the balance of payments. - Application of the direction principle is under study in a joint ECB-Eurostat working party; pending conclusions, some corrections made from January 1, 2003 for general direct respondents, and a change in data collection is programmed for 2006 to apply the direction principle fully while limiting respondent burden. ### 2.2 Scope - Coverage and residency - In principle, France’s balance of payments covers all transactions between residents and nonresidents. - BdF instructs respondents on the residency principle for completing payment reports (CRPs), statistical reports to direct reporters, and questionnaires; individual residence refers to primary center of interest (except official representative offices); legal entity residence is based on location of establishment and existence of real economic activity by autonomous production units. - Overseas Territories (legally not EU or euro-area members) are excluded from the monthly statistical coverage and from data forwarded to the ECB and Eurostat; their trade is reincorporated into quarterly and annual balance of payments figures sent to the IMF. ### 2.3 Classification/sectorization - Consistency with BPM5 - Institutional sector classification is consistent with BPM5; transaction classification is very close to BPM5. - “Travel” is broken down into professional and personal travel as recommended in BPM5, but only for annual data conveyed to Eurostat; breakdown uses ratios and could be made more regular based on survey data following a new approach introduced in 2002 (see 3.2.2). - c.i.f./f.o.b. adjustments for merchandise trade are based on survey data renewed every other year. - Estimates for passenger transportation portion of all-inclusive trips (transportation services) use ratios pending results of surveys introduced in 2003 (see 3.2.2). - Major construction and public works projects that under some BPM5 conditions could be direct investment are recorded as services due to absence of complete host-country activity accounts (see 4.3.3). - Insurance services are treated per BPM5: only the “net” portion of insurance services is recorded under services, with corresponding entries under current transfers; reinsurance services treated per BPM5 though no item for imputed household income from net equity in life insurance reserves. - Repos and reverse repos are treated as guaranteed loans, in accordance with BPM5. ### 2.4 Basis for recording - Valuation at market prices - Transactions on a cash basis correspond to market prices per BPM5. - Foreign exchange transactions converted at the average market exchange rate for the period. - When flows are derived from stock data, change in stocks in the settlement currency is valued at the average currency rate for the period; international investment position stocks are valued at end-of-period rate. - For mergers/acquisitions via public stock-for-stock offers, amount recorded is number of securities exchanged multiplied by stock market rate at the end of the offer; amount transmitted to reporting institution for verification. - International investment position: direct investment stocks are valued and reported both at book value and at market price; valuation of capital and reserves stocks depends on whether shares are listed or not and whether they are French or foreign. - Portfolio investment stocks are valued at market value; other investments recorded at book value except derivatives. - Security-by-security adjustments are possible due to detailed securities database at the DBDP (see 3.2.2). - Accrual recording practices and adjustments - DGDDI international trade data replace cash-basis ITRS data to be closer to BPM5 transaction concepts. - Timing adjustments made to record exports of refined petroleum on an accrual basis (at time product is sold rather than border crossing). - Income from debt securities reported to DBDP on a cash basis; pending conclusions of a European working group on accruals (DBDP participating), adjustments are made at time of transactions to show accrued interest under income by deducting it from settlement value and adding as positive under portfolio investment income. - These adjustments result in overestimation of income from debt securities for two transaction types: sales of French securities to nonresidents and purchases of foreign securities by residents. - Recording of debt forgiveness and rescheduling follows BPM5 (entries under capital transfers and appropriate financial account instrument); accumulation of payment arrears does not currently result in income entries with counterpart under other assets of general government, short term (arrears)—to be corrected when balance of payments and international investment position are calibrated against government accounts in the course of 2003. France is essentially a creditor country in these operations; impact on balance of payments is marginal. - Transfers from the European Community (entries under credits) are recorded on a cash basis; EAGGF transfers (Regulation No. 25 of 1962, amended most recently by Regulation CEE No. 728/70) are paid once a year but accumulated throughout the year as expenditure is incurred. - Income from other investments is recorded on a cash basis. - Grossing/netting procedures - In principle, current and capital transactions are recorded gross and financial account transactions are recorded net with asset and liability components broken down separately (BPM5 recommendation). - BdF adopted the new functional presentation of financial derivatives on January 1, 2000, using the option of presenting these transactions without distinguishing between assets and liabilities. *Source: _cr03339 - 1. Integrity* ### 3. Accuracy and reliability ### 3. Accuracy and reliability ### 3.1 Source data - The French balance of payments data collection is an international transactions reporting system (ITRS) relying on five principal sources: - (1) financial and monetary institutions reporting own transactions and those conducted on behalf of customers; - (2) the monetary authorities sector (operations of the Banque de France); - (3) general government for its own operations; - (4) major nonfinancial corporations covered by general direct reporting (DDG); - (5) other individuals and legal entities with accounts abroad covered by partial direct reporting (DDP). - Reserve assets data are collected from the Back-Office Directorate of the BdF. - Additional sources used to move from settlements statement to balance of payments statistics: - (1) DGDDI merchandise trade data collected from enterprises; - (2) monthly survey on trade credit and advances on orders (survey E84). - Other annual surveys complement data for the IIP and flow–stock comparisons; databases used for validation and market valuation include BAFI, FIBEN, and the security-by-security database. - Coverage and reporting population (as of end-December 2001): - about 350 banks with transactions abroad; - approximately 760 DDGs; - 1,800 DDPs. - Monthly transaction processing volume: - the DBDP receives, verifies, and analyzes between 600,000 and 800,000 international transactions each month. - Volume reduced after 1/1/2002 due to the European directive exemption threshold of €12,500 per operation. - Reporting rules and features: - Once the €12,500 threshold is exceeded, all transactions must be reported and classified by economic code. - Transactions are reported gross; subtraction of credits from debits is not authorized. - Each transaction produces a payment report (CRP) including the enterprise SIREN identifier. - Procedures for updating reporting bases: - DDG population: in theory updated once a year on final figures, but annual updating was not applied after 1997; planned reintroduction when new system for direct reporting on services is introduced in 2004. - DDP population: updated by comparing with bank declarations under code 062. - Populations for surveys E90 and direct investment stock surveys: updated annually on final flow figures. - MFI sector reporting population: updated annually. - Supplementary survey coverage: - Survey E84 (trade credits) covers 1,800 respondents, including 760 DDGs and about 1,100 SMEs. - Annual direct investment stocks survey covers about 2,000 direct investment enterprises in France and 9,000 enterprises for direct investment abroad. - Impact of €12,500 threshold on services estimates (loss in value estimated): - 8.7 percent on the revenue side (services overall); - 9.9 percent on the expenditure side (services overall); - For commercial services: 50.5 percent loss on the revenue side and 37.3 percent loss on the expenditure side. - A statistical procedure has been introduced to remedy these losses (see 3.2.2). - Monitoring and outreach: - Financial press monitored systematically by supervisory staff. - DDG enterprise websites monitored at least once a month by staff checking declarations. - Periodic meetings with professional organizations (e.g., AFTE) and with reporting firms; number of meetings with DDGs: 37 in 1998, 26 in 1999, 25 in 2000, 20 in 2001, and 25 in 2002. - Timeliness and deadlines: - General direct reporting monthly declarations: received not later than 30 days after month end; annual declarations by March 15 of the year following closing date of December 31. - Partial direct reporting declarations: received not later than the 15th of the following month. - Survey on stocks of French direct investment abroad: deadlines indicated on form; received by the deadline in over 90 percent of cases on average. - DBDP monitoring: weekly stock-taking, status reports, telephone/fax/email reminders. - 2002 response rates for settlements reported by DDGs: - about 80 percent gathered in time for production of the balance of payments for Key Items (m+35); - almost 95 percent by m+58. - Missing or invalid submissions are provisioned based on prior-month information; corrections recorded in statistical processing chains. ### 3.2 Statistical techniques - Data collection forms and processing: - Forms designed for enterprise ease and computer processing; most tested with enterprises. - SAFIR application adapted to efficient processing and allows numerous computerized checks. - Use of statistical models and supplementary surveys for compilation and estimation: - Estimation of unreported small amounts (below €12,500 threshold); - Estimation of wages of cross-border workers; - Calculation of accrued interest and reinvested earnings; - Estimation of direct investment stocks (survey virtually exhaustive); - Trade credits survey and two statistical surveys being tested to estimate tourism revenue and expenditure. - Procedure to estimate amounts missing due to the €12,500 exemption (post-January 2002): - Missing flows estimated on the basis of 2000 and 2001 figures. - Share of missing flows estimated for each balance of payments line using a study of monthly distribution in 2000 and 2001 of operations below the prior simplification threshold of €15,000 (used until 12/31/2001). - Statistical correction to continue in 2003 and subsequent years. - With the new system for general direct reporters planned for early 2004, a sample-based survey of non-DDG enterprises engaged in international services transactions is planned; by 2006 the new system is expected to largely compensate for information loss from the exemption threshold. - Adjustments to preliminary international trade data: - A statistical model adjusts preliminary DGDDI data to reduce revisions and support monthly balance of payments compilation at Key Items (m+35); DGDDI revisions spread over a 5-6 month period on average. - Corrections for missing MFI declarations: - If an establishment exceeding the €100 million threshold fails to submit required statements (10/11 or 12/13), the BAFI document (8010) is used, or the previous month's stock is carried forward. - Reinvested earnings estimation: - Initially estimated using a statistical model with assumptions based on INSEE’s half-yearly business cycle forecasts; about fifteen months after year end, calculated from accounting data on profits/losses and allocation accounts. - Provisional estimates calculated separately for resident enterprises with investment from abroad and nonresident direct investment enterprises. - INSEE forecasts used by industry, based on gross operating surplus from national accounts. - Final estimates for reinvested earnings of foreign branches of French enterprises based on profit/loss and dividend data collected via BdF branches and offices. - Final estimates for reinvested earnings of French branches of foreign firms use data from the enterprise banking file (FIBEN). - Tourism ("travel" line) estimation until completion of two surveys: - Pending surveys on residents’ expenditure abroad (April 2003) and nonresidents’ expenditure in France (September 2003), travel figures estimated from declarations by financial intermediaries, DDGs, and credit-card issuers, accounting for the share of credit-card spending in travelers’ payments. - Beginning in 2002 additional information identifies business vs personal credit-card purchases and isolates ATM withdrawals, Internet payments, and other payments. - After the 2003 surveys, provisional m+35 balance-of-payments data will be adapted to incorporate survey results and processing lags. - Valuation methods for stocks: - Stock of French listed direct investment companies valued using each enterprise’s share price; same method for listed foreign firms. - Nonlisted French firms: equity capital gathered on book value; valuation ratios (stock market value of listed firms/equity capital of listed firms) applied to equity capital of nonlisted firms. - Nonlisted foreign firms: cumulative valued flow method using annual balance sheet flows and market-value indexation via a composite foreign-stock-market index weighted by stock of French direct investment abroad and exchange rates. - Portfolio investment provisional international investment position uses a cumulative flow method with valuation indexes (French bonds, foreign bonds, share indexes, UCITS) developed by other BdF statistical units to ensure consistency with national financial accounts. ### 3.3 Assessment and validation of source data - Verification and control procedures: - Exhaustive verification processes at Direct Reporting (general and partial) levels. - Rapid manual checks for hard-copy or diskette submissions to identify and correct obvious errors. - SAFIR application provides three kinds of computerized controls: - (1) “Formal” controls (e.g., verification that countries, currencies, economic codes are in the collection nomenclature); - (2) Automatic consistency checks (CCAs) to identify largest operations and large changes in stocks (different thresholds depending on operation type), triggering in-depth examination and possible enterprise queries; - (3) Consistency checks on request (CCDs) for deeper examination of a respondent’s declaration (e.g., changes in stocks or duplication risks). - Time-series checks via status reports per enterprise recording two-year changes by code. - Bank reporting system checks: - Bank reporting structure consistent with bank accounting practices. - Balance of payments compilation not derived directly from BAFI statements but from specific balance-of-payments declarations. - Systematic cross-checks between balance-of-payments statements and BAFI statements for banks’ own-account operations; discrepancies prompt queries to banks. - Direct investment stocks survey checks: - Computer consistency check compares declared stock for year n with theoretical stock (stock n-1 plus flows in year n); sizable differences result in queries to the enterprise. - DBDP developed verification processes to identify errors in bank-reported data. - Time series by type of service systematically analyzed to identify anomalies, with monitoring of operations by enterprise. *Source: _cr03339 - 3. Accuracy and reliability* ### 3.4 Assessment and validation of intermediate data and statistical outputs ### 3.4 Assessment and validation of intermediate data and statistical outputs ### Validation of main intermediate data (3.4.1) - Revenue flows: - Time series are analyzed by type of revenue by cross-checking against revenue anticipated on the basis of estimated positions for the major securities lines (interest yields on public debt instruments, dividend payments). - Direct investment flows: - Completion of operations announced in the press is monitored. - The DBDP maintains a detailed list of operations announced in the financial press and forwards information to staff responsible for the enterprises concerned. - Portfolio flow declarations show data security by security, allowing pinpoint checks. - Issues announced in the press are monitored, as are redemptions. - Data on external public debt are compared with the government’s survey of “primary dealers.” ### Assessment and investigation of statistical discrepancies in intermediate data (3.4.2) - Corrections for lack of information or temporarily incomplete information: - For all BOP operations, corrections may be input to replace calculated data; these corrections are based upon expert analyses and validated at a plenary meeting with management. - Organizational changes: - The 2002 reorganization of DBDP units clarified the distinction between source data collection/production/validation and validation of statistical output, enabling greater emphasis on analyzing consistency over time and validation/justification of observed trends. - Presentation and cross-checks at validation meetings: - Source data are always presented in time series and checked at monthly/quarterly validation meetings against flows recorded in preceding months, cumulative year-to-date flows, annual flows, etc. - Detail of presentation varies by operation: - MFI operations: broken down by short- and long-term, by currency, by counterpart sector, and intra/extra breakdown. - Portfolio investment flows: broken down by category of securities (shares and UCITS, bonds and the equivalent, money market instruments). - Treasury securities: granular presentation for OAT, BTAN, BTF to closely monitor trading. - Other investments: lending and investment operations presented for four sectors (monetary authorities, general government, MFIs, other sectors); “other sectors” further distinguished by maturity (short-term vs. long-term), short-term investment operations (claims and liabilities), and assets of direct reporters with unaffiliated nonresidents. - Additional control tools: longer series such as monthly deposit/lending position of the MFIs, series on geographical distribution of direct investment, series on portfolio investments of BAFI reporting banks. ### Investigation of statistical discrepancies and other indicators of problems (3.4.3) - Cross-border wages: - European bilateral discrepancy analysis identified differences between French flows and nearby major trading partners, and a bias relative to BPM5 recommendations: settlements produced net flows while gross flows should be recorded. - As a result, flows of wages paid to cross-border workers residing in France and vice versa are now recorded gross for major trading partner countries bordering on France. - Impacts extended to travel (consumption of a portion of wages outside French territory) and current transfers (social welfare payments). - Additional cross-checks: - Quarterly calculations of implicit interest rate by major category of financial investment (credit and debit sides) by cross-checking against interest flows on stocks of corresponding financial assets. - Travel results compared with monthly survey of hotel occupancy conducted by INSEE on behalf of the Directorate of Tourism. - Study of international operations by enterprises in trade and services allows comparison with information from professional bodies by major sector. - European and bilateral work: - Numerous studies conducted to improve bilateral flows with France’s major trading partners and for compilation of the balance of payments of the Monetary Union; intra-European asymmetries remain a constant issue. - Preparation of a balance of payments for the Monetary Union led to use of national customs data in an “origin” methodology. - BdF adopted geographic breakdown of revenue flows from securities portfolio assets by place of issue. ### Revision studies (3.5) - Routine revision analysis (3.5.1): - Since 1992, DBDP prepares monthly results for deadlines varying between 40 and 65 days. - Increased publication frequency and shorter deadlines increased emphasis on measuring revisions between various values of balance of payments aggregates for the same period depending on status and production date. - Revision studies over the past ten years have been central to quality assurance; content of revisions at each deadline is analyzed with same frequency as publications. - Simple systematic analyses are conducted of revisions between the (three or four) published versions for the same month. - Some studies may disappear within the formalized framework for quality indicators defined at the European level. ### Serviceability — Relevance and user engagement (4.1) - User consultation and coordination: - DBDP is automatic member of CNIS "Money, Finance, and Balance of Payments" group; group meets twice a year to discuss statistical programs and gather respondent and producer opinions. - ECB staff produce periodic reports on quality of aggregates for the euro area and member country contributions. - Proposed statistical programs are discussed in European working groups and committees (e.g., Eurostat balance of payments working group, CMFB). - Regular meetings with major public sector users: - Half-yearly bilateral meetings with INSEE. - Monthly and half-yearly meetings with the Treasury (half-yearly: methodological/technical; monthly: macroeconomic/financial and includes other directorates). - DBDP participates in national accounts commissions and specialized commissions on services, tourism, and transportation. - Reports from meetings are published in the Journal officiel. - BdF represented on the IMF’s Balance of Payments Statistics Committee. ### Timeliness and periodicity (4.2) - Timeliness (4.2.1): - Dissemination deadline is 35 days following the end of the month for balance of payments Key Items, as required by the ECB, and shorter than SDDS timeliness of no more than one quarter after the end of the reference period. - Periodicity (4.2.2): - Detailed balance of payments is available monthly, exceeding SDDS quarterly requirement and meeting ECB requirements. ### Consistency (4.3) - Internal dataset consistency (4.3.1): - In France’s monthly settlements system, quarterly flows equal the sum of monthly flows and annual flows equal the sum of quarterly/monthly flows, except for financial flows derived from banking sector stocks where calibration accounts for valuation issues. - Time consistency (4.3.2): - Monthly data compiled in accordance with BPM5 available beginning in 1997; annual data beginning in 1995. - Annual series with a different classification available for 1973 to 1994. - Longer annual series available since 1946 on a basis comparable to but less detailed than BPM5. - Reconciliation with other data sources (4.3.3): - Since 1995 national accounts base, annual analysis conducted of discrepancy between balance of payments figures and rest-of-the-world account; cross-checking between INSEE and DBDP reduced nonmethodological discrepancies, notably for transportation services. - INSEE and DBDP working group established in late 2002 to aim at eliminating remaining discrepancies. - Noted methodological differences: - Major construction projects abroad: treated as construction services in French balance of payments but as direct investment in national accounts. - Transfers from the European Union: recorded on an accrual basis by national accounts and on a cash basis in the balance of payments. - Harmonization: - National financial accounts and balance of payments/IIP standards harmonized in context of ESA 95; a bridge table prepared jointly and valuation indexes used jointly to ensure consistency. - Eurosystem cross-cutting studies and comparisons (e.g., ECB comparisons of other investment of other sectors with mirror series sent to BIS). - Banking sector consistency: - Banking sector transactions in BOP consistent with money and banking statistics; consistency check with monetary statistics is a management rule for MFI sector data validation. - For “other investment—deposits/lending of MFIs,” systematic cross-checking of Key Items for month m, quarterly version, and annual version with monetary statistics in terms of stocks and flows. - Working group on “Counterpart/external position of MFIs” to report in 2003. - Identified methodological gaps for deposit/lending activity and MFI securities: (1) calculation of flows and valuation effects associated with exchange rate changes not performed exactly the same way; (2) revision procedures and timetable differ due to ECB policy differences; (3) procedures for correcting monthly data are not absolutely identical. ### Revision policy and practice (4.4) - Regular, transparent revision schedule (4.4.1): - Since 2002, revision policy: 1. publication of the initial, simplified monthly data (Key Items) at m+35. 2. publication of the following two months (in Key Items) without revision of the preceding months, at m+70 and m+105; 3. publication of data for the quarter at q+75 (data available with monthly periodicity and a detailed breakdown); 4. upon publication at q+75, only the months within the quarter in question are subject to revision; 5. publication of annual data at y+180 (data available with monthly periodicity and a detailed breakdown); 6. upon publication of the annual data, all months within the year in question are subject to revision; once these figures have been revised, the figures are final. - WD-SA (working days and seasonal adjustments) correction factors: - Re-estimated when data are published at q+75, using correction factors projected for data at m+35. - This minimizes changes in correction factors, particularly for working day corrections. - Only current account transactions are adjusted in this manner. - The revision policy is explained in a methodological paper available at no charge on the website. - Identification of preliminary data (4.4.2): - Statistical tables label data at m+35 as “provisional figures,” data at q+75 as “semifinal figures,” and data at y+180 as “final figures.” - Public availability of revision studies (4.4.3): - Methodological note “The balance of payments and the international investment position” explains sequence prompting the new revision policy and elimination of "provisional monthly results" published at m+60. - For current transaction flows, discrepancies over the 24 months of 2000 and 2001 between data at m+35 and m+60 were less than 1.6 percent on average on both revenue and expenditure sides. - Few revision studies are made available to the public. ### Accessibility — Data presentation and dissemination (5.1) - Presentation for interpretation (5.1.1): - BdF Bulletin (monthly) provides a brief two-page analysis of salient balance of payments developments with a simplified balance of payments table and charts based on moving 12-month WD-SA data. - Annexed statistical tables provide rapid reference; monthly statistics published in conformity with ECB Key Items for current account and a more detailed breakdown for financial account. - WD-SA data available for the six current account headings. - Annual Report on the Balance of Payments and the External Position of France includes extremely detailed analysis of developments. - Dissemination media and formats (5.1.2): - MINEFI and BdF press release is first information source; BdF publishes an analytical memorandum and updates databases on its website shortly after. - Bulletin and Annual Report available in hard copy or from BdF website; methodological notes and numerous time series provided free of charge on the BdF website, including pre-announced release schedule. - Latest BOP data and most recent Annual Report are available on MINEFI’s website. - Adherence to pre-announced schedule (5.1.3): - Data dissemination follows pre-announced schedule on BdF website indicating release dates for next four monthly sets. - Simultaneous release to all users (5.1.4): - Data released to all users at the same time, first through MINEFI and BdF press release on the MINEFI website and to press agencies, and rapidly thereafter on the BdF website. - Availability of nonpublished subaggregates on request (5.1.5): - DBDP can respond to requests for extremely detailed data subject to CNIS confidentiality rules (see 0.1.3). - BdF Bulletin and the balance of payments publications webpage provide contact details and e-mail address of a correspondent for additional information requests. *Source: _cr03339 - 3.4 Assessment and validation of intermediate data and statistical outputs* ### 5.2 Metadata accessibility ### 5.2 Metadata accessibility ### Documentation and methodological transparency - A detailed methodology is published in the Annual Report and is accessible via the BdF website using a navigation tool. - The Annual Report contains methodological text boxes that cover specific topics in detail (see 4.3.2). - Articles on methodology are published in the Bank’s Bulletin (e.g., an article on the balance of payments methodology of the euro area appeared in the Bulletin for November 2002). - A more comprehensive methodological note explaining principles, compilation stages, structure and balances of the balance of payments and the IIP, and work in progress and changes planned for 2003, was published on the BdF website in December 2002. - The chapter on France in the ECB’s “BOP Book,” accessible on the ECB website, reproduces the DBDP methodology in accordance with the ECB format required. ### Levels of detail for audiences - Different publication levels address varied audience needs: - Simplified analytical presentation in the Bulletin. - Exhaustive treatment in the Annual Report. - Statistical tables and methodological notes at greater detail for technical users. ### Assistance to users and user contact - Two three-person units service inquiries: one for institutional users and one for the general public. Contact channels (telephone, fax, e-mail) are widely disseminated and included in hard-copy and web publications from the DBDP. - Publication tables and charts (Bulletin, Annual Report) cite the data source and unit responsible, including a telephone number. - The DBDP maintains a detailed log of requests received during the current year. In 2002, the DBDP responded to 180 individual data requests. - Requester composition in 2002: - General government: 26 percent - Embassies: 24 percent - Subject focus in 2002: - Direct investment: 59 percent - Interest in methodology and the information notice: roughly 10 percent ### Catalogues and dissemination services - A catalogue of BdF publications, indicating the cost of the products, is published annually by the Communications Directorate of the BdF and is accessible on the BdF website. - The Public Relations Unit of the Communications Directorate, together with “Administration, Dissemination, Micro and Information System,” services orders. ### Data quality assessment: key observations (Table 6 summary) - Prerequisites of quality: - Comment: The MINEFI is identified as the source of the statistical data it disseminates, which may give the impression that it has been involved in preparing the figures. The statistical production of the balance of payments could be better integrated in the French statistical system. - Methodological soundness: - Comment on basis for recording: The adjustments for accrued interest in respect of bond transactions overestimate gross income. - Serviceability / Consistency: - Comment: There are differences between the current account of the balance of payments and the rest-of-the-world sector of the national accounts. - Overall table coding uses NA / O / LO / LNO / NO classifications across elements (as presented in the source). ### SDDS and metadata practices (Appendix I highlights) - SDDS prescribes: - Dissemination of 18 data categories with component detail covering the four main sectors of the economy, with prescribed periodicity and timeliness. - Advance release calendars with at least one-quarter advance notice of approximate dates and at least a one-week advance notice of precise dates; and simultaneous release of data to all users. - Dissemination of terms and conditions under which official statistics are produced; identification of internal government access to data before release; identification of ministerial commentary on statistical release; provision of information about revisions and advance notice of major methodological changes. - Dissemination of documentation on statistical methodology and sources; dissemination of component detail/additional series to permit cross-checks. - SDDS subscribers undertake to post descriptions of data dissemination practices (metadata) on the IMF’s DSBB and to maintain a National Summary Data Page linked electronically to the DSBB. - IMF Executive Board approvals and monitoring timing noted: - At the March 29, 2000 meeting, a new SDDS data category on external debt was incorporated; the transition period for implementing this data category expires in March 2003. - Monitoring of observance by IMF staff commenced at the beginning of July 2000 and is limited to coverage, periodicity, timeliness, and dissemination of advance release calendars. - Reference cited in source: Guide to the SDDS, February 2002: http://dsbb.imf.org ### User feedback from the informal survey and meeting (Appendix III) - Survey logistics and response: - INSEE identified 115 data users; 46 questionnaires were returned (response rate 40 percent). - A meeting of about ten users was held on March 7, 2003; attendants confirmed questionnaire comments. - General findings: - Users regard France’s macroeconomic statistics as overall above the average for other European countries; national accounts viewed as far above the average. - Coverage is considered quite satisfactory overall except for government finance statistics; level of detail is satisfactory except for government finance statistics and the balance of payments. - Periodicity satisfactory except for government finance statistics and to some extent monetary statistics. Timeliness satisfactory except for national accounts. - Users aware of release calendars for national accounts and price indexes but less so for other macroeconomic statistics; when aware, pre-announced dates are generally observed. - A significant proportion of users considers that revisions are not sufficiently documented. - Access to data and to metadata is regarded as satisfactory overall; metadata are considered sufficiently clear and detailed. - Methodologies appear sound and appropriate; accuracy and impartiality ranked highly. - Areas for improvement emphasized: government finance statistics and social statistics; businesses and federations find unmet needs; accessibility to long series should be facilitated; clearer presentation of interconnections between statistical fields. - Principal user comments and suggested improvements (selected) - Insufficient coverage: - Extend production and consumption statistics to cover the services sector. - Strengthen coverage of general government beyond central government (local governments, hospital sector, various central government agencies, social security). - Indicate the real coverage of the balance of payments by comparison with its theoretical coverage. - Add information on stocks and turnover. - Insufficient level of detail: - Government finance statistics should distinguish local taxation from national taxation, and taxation of labor from taxation of capital. - Breakdown of public expenditure by function; more granular information on total public debt. - Monetary statistics: more detailed breakdown by purpose of loans and by average maturity. - More enterprise-size detail and more detailed breakdown of operations with the rest of the world (separate out the euro area and, if possible, by country). - Periodicity and timeliness: - Quarterly accounts sometimes take too long to appear; France and other European countries often slower than the United States. - Revision practices: - INSEE documents revisions well but notification should be more prominent. - Backward extrapolation of series should maintain original level of detail. - Release figures more rapidly even if later revision is necessary; revised series are published too late. - Access to methodologies and data: - National accounts: documentation on methods inadequate. - Not enough documentation online; methodologies should be written in more readily comprehensible terms for the general public. - Improve search engines for downloadable series; allow downloading of groups of series; facilitate access to long series. - Government finance statistics: data dispersed across multiple ministry/administration websites; lack transparency and clarity. - Quality of methodologies and comparability: - Methodologies cohesive but often display discrepancies vis-à-vis other national or European standards and nomenclatures. - Price indexes: doubts about weighting. - Government finance statistics: complex interconnections with national accounts and the government budget. - National accounts: problems linking nomenclatures for price, volume, value, employment, and capital amounts. - Accuracy and impartiality: - National accounts perceived by some users as too optimistic. - Areas for enhancement: - Strengthen social statistics, labor market, and wages in methodology, periodicity, and accessibility. - Address data needs of extremely small businesses and crafts. *Source: IMF staff mission report content (sections 5.2–5.3, Table 6, Appendices I–III as provided in the supplied document).* --- _Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2003/_cr03339.pdf_