## _cr07318

## Source details

**Canonical URL:** [_cr07318](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07318.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07318.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07318.pdf.json)

---

### Overall assessment
- Chile subscribed to the SDDS on May 17, 1996 and started posting metadata on the DSBB on February 19, 1997.
- Chile is in observance of the SDDS; uses a flexibility option on timeliness of central government operations data (currently meets SDDS requirements).
- This ROSC data module reassesses April 2001 exercise using the IMF’s DQAF (July 2003) for NA, prices (CPI, PPI), GFS, monetary, and BOP statistics.
- Key strengths:
  - Macroeconomic statistics timely and generally of high quality.
  - High degree of quality awareness and reputation of integrity.
  - Major methodological improvements since 2001: change of NA benchmark year; compilation of a PPI; implementation of BPM5; adoption of GFSM 2001.

### Prerequisites of quality — institutional responsibilities and resource findings
- Institutional responsibilities:
  - NSI: price statistics.
  - CBCH: national accounts, monetary, and balance of payments statistics.
  - DIPRES (MOF): government finance statistics.
  - SBFI: collects monetary data.
- Legal / institutional needs:
  - Independence of NSI needs reinforcement (director/deputy tenure: three years plus one possible renewal; discretion for removal).
  - No general legal precept empowering CBCH to require private institutions to respond solely for statistical purposes (Article 40 empowers CBCH to demand information on foreign exchange transactions).
- Resource assessments (verbatim):
  - Resources adequate at CBCH and at SBFI.
  - Resources need strengthening at DIPRES (MOF) to complete migration to GFSM 2001 and ensure business continuity and data sharing.
  - Resources inadequate at NSI to ensure:
    - regular updating of the business register;
    - well-diversified outlet and enterprise samples;
    - continuous methodological enhancements and validation/timeliness of source data for NA and price statistics;
    - reliable business continuity planning;
    - appropriate quality adjustments to price statistics.
- Confidentiality and access:
  - Law 17374/1970 guarantees confidentiality (Art. 29 and 30); CBCH confidentiality guaranteed by Article 66 and By-Laws.
  - Some internal pre-release access vulnerabilities noted (e.g., Corporate Image Department access to CPI/PPI results).

### Methodological soundness — major dataset findings and needs

- National Accounts (NA)
  - Concepts/standards: compiled following the 1993 SNA.
  - Scope: annual and quarterly GDP at current and constant 2003 prices; annual SUT; I-O matrix for base year; institutional sector accounts to financial account.
  - Deviations / areas for improvement:
    - Treatment exceptions: mineral exploration, agricultural work in progress, own-account production of software/databases/manufacturing products.
    - Classification: copper derivatives classification deviates from ISIC Rev. 3; COFOG and COICOP not yet used for government and household final consumption classifications.
    - Source data issues: NSI business register partially updated; HBS conducted every 10 years instead of five.
    - Statistical techniques: double-deflation used for around half of value added; changes in inventories obtained as residuals; Denton method for benchmarking; quarterly data seasonally adjusted.
  - Dissemination and metadata:
    - Preliminary GDP meets SDDS periodicity/timeliness: quarterly timeliness examples 7W; 11W for Q4.
    - Detailed metadata available in CBCH special publications; SDDS metadata need updating.

- Consumer Price Index (CPI)
  - Concepts: broad correspondence with CPI Manual, 2004.
  - Scope: fixed-basket representative of households of Greater Santiago (about 40 percent of Chile’s population); owner-occupied housing not currently included.
  - Major issues:
    - CPI basket includes interest rates on consumption loans and mortgage payments with inconsistent treatment (combined weight about 3.6 percent).
    - Geographic coverage limited; CPI basket and weights based on 1996–1997 HBS.
    - Erroneous treatment of used cars weight (used cars weight 1.12 percent; new cars weight 0.67 percent).
    - Classification: education split into different product payments (education services weight 6.2 percent) — potential downward bias.
    - Quality adjustments generally not made (exception: computers, procedure requires revision).
  - Timeliness and dissemination:
    - CPI published monthly within five working days after month end (meets SDDS).
    - Press releases lack appropriate analysis and core measures presentation; NSI does not publish CBCH’s X1 core inflation index.

- Producer Price Index (PPI)
  - Scope: covers six categories (agriculture, livestock and forestry; mining; fishing; manufacturing; electricity, gas and water; construction); does not cover services. Monthly collection: about 4,000 goods in 738 establishments.
  - Key shortcomings:
    - Does not record copper price movements on an accrual basis; copper PPI weight 13.3 percent.
    - Adjustments for quality differences not made; missing prices are often carried forward.
    - Public knowledge and credibility low; PPI metadata limited; detailed PPI time series not readily available on NSI website (historical series action taken in responses).
  - Publication: monthly within five working days (meets SDDS); an indicator excluding copper effects not published (pilot tests exist).

- Government Finance Statistics (GFS)
  - Concepts: broadly consistent with GFSM 2001; DIPRES documented migration plan since 2003.
  - Institutional/transactional coverage incomplete: universities and ancillary units serving local governments not covered.
  - Deviations from GFSM 2001:
    - Central government debt excludes accrued interest on debt instruments, liabilities related to Reserved Copper Law, stock of recognition bonds (supplementary data disseminated separately).
    - Tax revenues recorded on a cash basis.
    - Interest on indexed debt instruments does not include changes in principal due to index fluctuations.
  - Data/compilation:
    - SIGFE provides comprehensive accounting for budgetary central government; stock data recorded at face value.
    - GFS not internally consistent due to differences in source data and valuation methods for debt flows and stocks.

- Monetary statistics
  - Broadly in line with MFSM; DC survey coverage complete (CBCH, commercial banks, Banco Estado, credit unions, mutual funds).
  - Three nonmaterial departures from MFSM:
    1. Instruments valued at lower of amortized cost or market prices rather than market prices.
    2. Interest not accrued on nonperforming loans (recorded in memorandum accounts).
    3. Gold holdings recorded using average of daily prices of last three months rather than closing day quotation.
  - Sectorization issues: securities issued by CBCH and liabilities of mutual funds not properly sectorized in some surveys.
  - Timeliness: monetary statistics disseminated on the 7th and 23rd day after reference month; provisional for two months then final.

- Balance of Payments (BOP)
  - Broadly follow BPM5; implementation completed in 2002 with series from 1996 onwards.
  - Exceptions in recording basis:
    - Loan liabilities of public and private sector recorded on a cash basis; portfolio and other investment income generally on cash basis for liabilities and accrual for assets.
    - Some interest on external debt recorded on a cash basis.
  - Source data: customs, CIER, surveys, administrative records; DBPDE uses robust techniques and computerized edit checks.
  - Timeliness: BOP and IIP meet SDDS; partial frequencies include weekly (external trade), semi-monthly (reserve assets), monthly (financial account flows).
  - Revisions: quarterly data revised each time a new quarter is released; annual revisions in March of the following two years.

### Accuracy and reliability — source-data and validation findings
- General assessment: overall accuracy and reliability adequate; source data and statistical techniques sound; outputs sufficiently portray reality.
- Common concerns across datasets:
  - Rapid expansion of services and financial system challenges source data and international-standard compliance.
  - Business register partially outdated and not regularly updated.
  - HBS frequency (every 10 years) not in line with best practice; affects household consumption expenditure data for NA and CPI weights.
  - Lack of explicit estimates of nonobserved activities (informal/unrecorded/underground).
  - For most surveys, scientific sampling techniques not used; HBS and many business surveys lack documented sampling/nonsampling error estimates.
  - PPI and CPI validation thresholds invariant across items (e.g., CPI uses uniform +/- 10 percent investigation threshold).
  - GFS stock data recorded at face value only; recommendation to value stocks at market or nearest equivalent.
  - Monetary statistics would benefit from additional cross-validations with fiscal data.

### Serviceability and accessibility — dissemination and user access
- Serviceability:
  - Periodicity and timeliness generally adequate; most datasets provide public revision schedules and identify provisional estimates.
  - Some interdataset inconsistencies exist (NA vs GFS; monetary net credit to government vs fiscal data; BOP vs external debt) but are reconcilable or noted.
  - GFS not internally consistent and not routinely reconciled with other agencies’ data (except NA).
- Accessibility:
  - Macroeconomic statistics readily available electronically; CBCH data particularly easily downloadable.
  - Price statistics presentation hinders interpretation and comparisons; press releases lack appropriate analysis.
  - NSI does not publish X1 price index though CBCH prepares it.
  - PPI indicator excluding copper effects is not published (pilot tests exist).

### Users’ views — survey and consultations (exact figures preserved)
- Users’ survey: questionnaires sent 207; responses received 50 (24.2 percent response rate).
  - By type (Total Sent / Total Received / In % Received): Financial institutions 45 / 20 / 44.4; Universities/researchers/consultants 94 / 14 / 14.9; Government 20 / 6 / 30.0; Media 29 / 1 / 3.5; Producers’ associations 4 / 4 / 100.0.
- Key user findings:
  - Most users satisfied with methodological soundness, timeliness, accessibility, and frequency.
  - Areas for improvement:
    - PPI credibility and usage limited.
    - Coverage/detail of GFS and PPI insufficient.
    - Limited awareness of early release calendar for GFS.
    - Information on revisions for NA, PPI, GFS, and BOP could be improved.
    - Demand for more detailed and longer time series on NA, BOP, PPI, IMACEC, and GFS.
    - Interest in household income distribution, average prices by industry, volume of sales by industry, and reliable employment series.
- Selected user-survey statistics (exact values):
  - Which statistics used regularly (Total Number of Responses): NA 38; Prices 48; GFS 35; Monetary 32; BOP 30; Other 38.
  - Access channels used: Official website 49; Official press releases/publications 46; Hard copy publications 30.
  - Overall quality rating (1 poor to 5 excellent) — Average rating: NA 3.9; Prices 4.0; GFS 3.7; Monetary 4.3; BOP 4.2.
  - Knowledge of a publicly disseminated release calendar: NA Yes 26 / No 18; GFS Yes 15 / No 27; Prices Yes 36 / No 13.

### Principal needs and action areas highlighted by the assessment (selected high-priority)
- Legal and institutional:
  - Reinforce NSI legal independence (lengthen tenure and reduce discretionary removal).
  - Provide legal authority for CBCH to require private-sector data for statistical purposes beyond foreign exchange transactions.
- Resources and capacity:
  - Strengthen NSI resources to update business register regularly; diversify outlet and enterprise samples; implement business continuity planning; apply quality adjustments to price statistics.
  - Strengthen DIPRES resources to complete migration to GFSM 2001 and ensure business continuity and data sharing.
- Interagency coordination:
  - Improve coordination to ensure intersectoral consistency and reconciliation across NA, GFS, monetary, and BOP statistics.
- Prices:
  - Enhance CPI geographic coverage; update CPI basket and weights (new HBS to establish a new CPI base period by December 2008).
  - Improve treatment of financial components and educational services classification.
  - Improve PPI credibility by publishing more detailed metadata and time series, addressing copper-related accrual recording and quality adjustment procedures, and publishing a PPI excluding copper effects.
- Source data and estimation:
  - Expand and modernize source data for agriculture, fishing, services, and quarterly estimates; increase use of scientific sampling techniques and estimate sampling/nonsampling errors.
  - Develop explicit estimates of nonobserved activities and improve inventory measurement in NA.
  - Increase cross-validation and reconciliations between monetary, fiscal, GFS, and external debt statistics.

### Cross-cutting and institutional recommendations (selected high priority; verbatim actions)
- Ensure NSI has adequate financial resources, staff, facilities, and training; increase retention of qualified staff.
  - NSI: considered an increase in the budget for next four years to create a CPI study department and retain qualified staff participating in new CPI 2008; request included in INE’s budget proposal.
- Strengthen independence of NSI by lengthening tenure from three to six years, staggering appointments, maintaining reappointment option, limiting discretion for removal.
  - Authorities: matter beyond INE’s authority; requires Law of Civil Service changes.
- Enhance coordination among statistical agencies via regular technical meetings; Committee INE-BCCH creation to revise classifiers/nomenclatures; first reconciliation results expected by end of the year.
- Reinforce CBCH legal framework to request data from nonfinancial private sector.
  - BCCH: Article 53 authorizes requesting information from public entities; no amendment planned; BCCH obtains satisfactory responses despite lack of explicit private-sector mandate.
- Initiate regular consultations with public and private users; foster user groups.
  - Authorities: INE creating a Committee of Users; CBCH to include formal user consultations beginning in 2008.
- Intensify intersectoral data consistency and reconciliation by addressing coverage, classification, and sectorization differences.

### Authorities’ responses and selected work program highlights (verbatim)
- Central Bank of Chile — National accounts (high priority):
  - Update business register regularly by requesting annual sales and employment by enterprise to ITS; conduct simplified economic census every five years.
  - 2008 benchmark compilation: examine statistical separation of copper activity between mining production and refined goods; increase detail of disseminated data at three digits of ISIC (or at least two digits).
  - Reconstruct historical series when methodology/source changes occur; 2008 work program includes review of statistical retropolation methods.
- CBCH — Monetary statistics:
  - Develop methodology to identify holders of CBCH securities and consult SVS regarding mutual fund liabilities identification.
  - Explore additional validations and compilations for other financial corporations (pension funds, insurance corporations, other intermediaries).
- CBCH — Balance of payments statistics (high priority):
  - Coordinate with ITS to reduce time to access tributary data to improve quarterly BOP estimations.
  - Design and implement a data warehouse for external statistics (BOP, IIP, external debt); bidding underway for external consulting.
  - Record external debt transactions on an accrual basis; CBCH intends to incorporate accrual recording into BOP and external debt statistics in March 2008, conditional on Statistics Committee approval, with revised figures for 2005 and 2006.
  - Record net creditor balance on ALADI reciprocal credits under Reserve Assets — BCCH notes matter numerically insignificant (less than 1 percent of reserves).
- NSI (INE) — Price statistics (selected actions):
  - Implement ISO 9001:2000 quality management system on CPI by December 2007 and on PPI by December 2008.
  - HBS: new national HBS underway; plan to update CPI base with expenditure results by December 2008; requesting resources to levy HBS every five years.
  - CPI/PPI methodological improvements: revise financial expenditure treatment; reclassify educational services; revise used cars weight methodology; expand geographic scope beyond Greater Santiago; publish detailed PPI historical time series and a PPI excluding copper.
  - Improve presentation and content of press release; develop more user-friendly webpage; publish contact persons in publications.
- DIPRES (MOF) — GFS actions:
  - Follow GFSM 2001 guidelines in classifying revenue, expenses, capital transfers, and financing transactions.
  - Improve institutional and data coverage of general government: include universities and ancillary units; record capitalized interest on fiscal promissory notes as interest expense; include accrued unpaid interest and recognition bonds in liabilities.
  - Reconcile debt stocks and flows periodically to improve horizontal consistency.
  - Aim to establish cooperation agreement with CBCH to formalize data sharing.

### SDDS coverage, periodicity, and timeliness — selected metadata (verbatim examples)
- National accounts: Periodicity Q; Timeliness Q; Chile timeliness details: 7W; 11W for Q4.
- Production index / Global production index: Periodicity M; Timeliness M; Chile timeliness details: 6W (1M encouraged); 4W; 35D.
- IMACEC / Forward-looking indicators: CC 2Q; IPE 36D.
- Employment / Unemployment: Periodicity Q; Timeliness Q; Chile timeliness details for unemployment: 4W.
- CPI: Periodicity M; Timeliness M; Chile timeliness details: 5D.
- PPI: Periodicity M; Timeliness M; Chile timeliness details: 5D.
- General government operations (Fiscal) (annual): Chile Timeliness: 2Q; Timeliness details: 6M.
- Central government operations (monthly): Chile avails flexibility option — Timeliness: 1M.
- Central government debt: Periodicity Q; Timeliness Q; Chile timeliness: 3M.
- Balance of payments: Periodicity Q; Timeliness Q; Chile timeliness details: 7W — Fourth quarter, 83 days; other quarters, 53 days.
- Official reserve assets: Periodicity M; Timeliness W; Chile timeliness examples: M 1W; 23 days.
- International Investment position: Periodicity A (Q encouraged); Timeliness: 6M; Chile periodicity A; Timeliness: 3Q (1Q encouraged); 2Q.
- External debt: Periodicity Q; Timeliness M; Chile timeliness: Q; M; Q; 5W.

*Source: DQAF July 2003 — Summary Results (ROSC data module assessment), and detailed DQAF assessments for NSI (INE), CBCH, DIPRES, SBFI, and associated dataset chapters as contained in _cr07318.*

### 1.      DQAF July 2003—Summary Results...............................................................................9

### 1.      DQAF July 2003—Summary Results

### Overall assessment
- Chile subscribed to the Special Data Dissemination Standard (SDDS) on May 17, 1996 and started posting its metadata on the Dissemination Standards Bulletin Board (DSBB) on February 19, 1997.
- Chile is in observance of the SDDS, meeting the specifications for coverage, periodicity, timeliness, and the dissemination of advance release calendars.
- Chile uses a flexibility option on the timeliness of the central government operations data; although at present it meets SDDS requirements in this regard.
- This ROSC data module is a reassessment of the exercise conducted in April 2001 and applies the IMF’s Data Quality Assessment Framework (DQAF July 2003) to national accounts (NA), prices, government finance statistics (GFS), monetary, and balance of payments (BOP) statistics.
- Institutional responsibilities:
  - National Statistics Institute (NSI): price statistics.
  - Central Bank of Chile (CBCH): national accounts, monetary, and balance of payments statistics.
  - Budget Directorate (DIPRES), Ministry of Finance (MOF): government finance statistics.
  - Superintendency of Banks and Financial Institutions (SBFI): collects monetary data.

### Key strengths
- Macroeconomic statistics are timely, generally of high quality, and adequate for effective surveillance.
- High degree of quality awareness among Chile’s statistical managers and a reputation of integrity of statistical institutions and processes among data users.
- Agencies demonstrate professionalism, transparency, and documented practices to ensure impartiality.
- Major methodological improvements since 2001 include:
  - Change of NA benchmark year.
  - Compilation of a producer price index (PPI).
  - Implementation of BPM5.
  - Adoption of GFSM 2001.

### Prerequisites of quality — main findings and needs
- Legal and institutional frameworks are generally adequate, but:
  - Independence of the NSI needs to be reinforced (short tenure of directors and deputy directors: three years plus one possible renewal, and discretion for removal).
  - No legal precepts empowering the CBCH to require private institutions to respond to data requests solely for statistical purposes, except for information on foreign exchange transactions (Article 40 of the CBCH’s Basic Constitutional Act allows it to demand information on foreign exchange transactions from the private sector).
- Resource assessments:
  - Resources are adequate at the CBCH and at the SBFI.
  - Resources need strengthening at DIPRES (MOF) to complete migration to GFSM 2001 and ensure business continuity and data sharing.
  - Resources are inadequate at the NSI to ensure:
    - regular updating of the business register;
    - well-diversified outlet and enterprise samples;
    - continuous methodological enhancements and validation/timeliness of source data for NA and price statistics;
    - reliable business continuity planning;
    - appropriate quality adjustments to price statistics.

### Assurances of integrity
- Agencies adhere to the principle of objectivity in collection, processing, and dissemination.
- Chile’s macroeconomic and financial institutional environment rests on four pillars:
  1. autonomous central bank with inflation targeting;
  2. fiscal policy governed by a fiscal rule aimed at achieving a structural fiscal surplus, considering long-run copper prices and projected copper output levels;
  3. regulation and supervision of the financial sector according to international standards involving several supervisory entities (including the SBFI);
  4. integration with international markets through trade openness, free capital mobility, and a floating exchange rate regime.
- Participation in international statistical fora and exercises enhances professionalism (examples cited: IMF’s Balance of Payments Committee, International Comparison Program, UN Ottawa Group on Price Indices, OECD observer status, and participation in the Coordinated Compilation Exercise for Financial Soundness Indicators).

### Methodological soundness — main findings and needs
- Macroeconomic statistics broadly follow internationally accepted standards, guidelines, and good practices.
- Areas for improvement:
  - National accounts scope: treatment of mineral exploration, agricultural work in progress, and own-account production of computer software, databases, and manufacturing products.
  - Classification and sectorization: classification of copper derivatives; functions of government; individual consumption by purpose (improvements under development).
  - CPI limitations:
    - Limited geographic coverage (plans to address; a new HBS with national coverage is being levied to establish a new CPI base period by December 2008).
    - Basket includes interest rates on consumption loans and mortgage payments with inconsistent treatment relative to best practice.
    - Erroneous classification of educational services may introduce a downward bias.
  - PPI issues:
    - Does not register copper price movements on an accrual basis.
    - Adjustments for quality differences are not made; missing prices are carried forward from the last observed price in most categories.
    - Public knowledge and credibility of the PPI are scant; PPI metadata on the NSI website are limited (comprehensive CPI metadata available; only summary PPI metadata and no detailed PPI time series).
  - GFS deviations from GFSM 2001: incomplete institutional and transactional coverage for general government; significant deviations in classification systems; aggregates not conforming to GFSM 2001 include revenues from copper, subsidies and grants, capital transfers, financing transactions, and the stock of central government debt.
  - Monetary statistics: broadly in line with MFSM, with nonmaterial departures (valuation of financial instruments, accrual of interest on nonperforming loans, recording of gold holdings). Securities issued by the CBCH and liabilities of mutual funds not properly sectorized in some surveys.
  - BOP: broadly in line with BPM5, but some departures where interest on external debt is recorded on a cash basis and some issues in recording certain FDI transactions.

### Accuracy and reliability — main findings and needs
- Overall accuracy and reliability are adequate; source data and statistical techniques are sound and outputs sufficiently portray reality.
- Concerns and limitations:
  - Rapid expansion of services and the financial system presents challenges to source data and compliance with international standards.
  - Business register is partially outdated and not updated regularly.
  - Frequency of the Household Budget Survey (HBS) is not in line with international best practice, potentially affecting household consumption expenditure data for NA.
  - Source data for agriculture, fishing, some services, and quarterly estimates are limited in scope.
  - Lack of timeliness in validating NSI survey data limits potential use.
  - For most surveys, scientific sampling techniques are not used; HBS and most business surveys lack information on sampling and nonsampling errors.
  - No explicit estimates of nonobserved activities (informal, unrecorded, and underground activities).
  - Changes in inventories are obtained as residuals from supply-and-use tables, resulting in erratic price movements.
  - CPI basket and weight structure are outdated (plans to address).
  - CPI source data consistency is not verified with other sources; procedures to refresh outlets in the CPI sample are not in place; adjustments for quality differences are generally not made; sources of errors are not routinely investigated.
  - PPI copper price recollection procedure increases quality-induced volatility.
  - GFS stock data on financial assets and liabilities are recorded at face value only.
  - Monetary statistics could be enhanced via additional validations and cross-checking with fiscal data.

### Serviceability and accessibility — main findings
- Serviceability:
  - Macroeconomic statistics have adequate periodicity and timeliness.
  - Most datasets provide public statements of revision schedules and identify provisional estimates.
  - Datasets are broadly consistent; some inconsistencies (NA vs GFS; monetary net credit to government vs fiscal data; BOP vs external debt statistics) are reconcilable or noted.
  - GFS are not internally consistent due to differences in source data and valuation methods for debt flows and stocks and are not routinely reconciled with other agencies’ data (except NA).
- Accessibility:
  - Macroeconomic statistics are readily available electronically and generally easily downloadable from data-producing agencies’ websites (particularly CBCH).
  - Price statistics presentation hinders interpretation and meaningful comparisons; press releases lack appropriate analysis of current-period price developments.
  - NSI does not publish the X1 price index (CBCH’s core inflation definition) though CBCH prepares the estimate.
  - A PPI indicator that excludes copper effects is not published.

### Users’ views (survey and consultations)
- A users’ survey with 50 respondents was conducted; meetings with selected users were held.
- Most users were satisfied with methodological soundness, timeliness, accessibility, and frequency.
- Areas identified by users for improvement:
  - PPI credibility and usage are limited.
  - Coverage and detail of GFS and PPI seen as insufficient.
  - Awareness of existence and observance of an early release calendar for GFS was limited.
  - Information on revisions for NA, PPI, GFS, and BOP could be improved.
  - Access to metadata on NA, PPI, and GFS could be improved.
  - Demand for more detailed and longer time series on NA, BOP, PPI, IMACEC, and GFS.
  - Interest in data on household income distribution, average prices by industry, volume of sales by industry, and reliable and consistent employment series.

### Assessment by agency and dataset
- The mission applied the DQAF July 2003 and presents results using a four-point rating scale (Table 1): O = Practice Observed; LO = Practice Largely Observed; LNO = Practice Largely Not Observed; NO = Practice Not Observed; NA = Not Applicable.
- Prerequisites of data quality and assurances of integrity (Dimensions “0” and “1”) are presented for agencies in Tables 2a–d.
- For each dataset, the assessment of methodological soundness, accuracy and reliability, serviceability, and accessibility (Dimensions “2” to “5”) are shown in Tables 3a–f.

### Principal needs and action areas highlighted by the assessment
- Reinforce legal independence and statutory protections for the NSI.
- Provide legal authority for CBCH to require private-sector data for statistical purposes beyond foreign exchange transactions.
- Strengthen NSI resources to:
  - update the business register regularly;
  - diversify outlet and enterprise samples;
  - carry out continuous methodological improvements and source-data validation;
  - implement business continuity planning;
  - apply quality adjustments to price statistics.
- Strengthen DIPRES resources to complete migration to GFSM 2001 and ensure business continuity and data sharing.
- Improve interagency coordination to ensure intersectoral consistency and reconciliation across NA, GFS, monetary, and BOP statistics.
- Enhance CPI geographic coverage, update CPI basket and weights (new HBS to establish a new CPI base period by December 2008), and improve treatment of financial components and educational services classification.
- Improve PPI credibility by publishing more detailed metadata and time series and by addressing copper-related accrual recording and quality adjustment procedures.
- Expand and modernize source data for agriculture, fishing, services, and quarterly estimates; increase use of scientific sampling techniques and estimate sampling/nonsampling errors where possible.
- Develop explicit estimates of nonobserved activities and improve inventory measurement in NA.
- Increase cross-validation and reconciliations between monetary, fiscal, GFS, and external debt statistics.

*Source: DQAF July 2003—Summary Results (ROSC data module assessment).*

### 0. Prerequisites of quality

### _cr07318 - 0. Prerequisites of quality

### Overview: assessment framework and practice categories
- Dimensions assessed: 0. Prerequisites of quality; 1. Assurances of integrity; 2. Methodological soundness; 3. Accuracy and reliability; 4. Serviceability; 5. Accessibility (table headings present in source).
- Practice categories (verbatim text from source):
  - Practice observed: current practices generally in observance meet or achieve the objectives of DQAF internationally accepted statistical practices without any significant deficiencies.
  - Practice largely observed: some departures, but these are not seen as sufficient to raise doubts about the authorities’ ability to observe the DQAF practices.
  - Practice largely not observed: significant departures and the authorities will need to take significant action to achieve observance.
  - Practice not observed: most DQAF practices are not met.
  - Not applicable: used only exceptionally when statistical practices do not apply to a country’s circumstances.

### National Statistics Institute (NSI) — Key findings
- Legal and institutional environment:
  - Law 17374 of 1970 guarantees the confidentiality of individual reporters’ data (breach of confidentiality has a criminal punishment, Art. 29 and 30).
  - The Law does not specify the responsibility of the NSI to compile source data for national accounts purposes.
  - Statistical reporting is mandatory according to the Law (penalties specified in Art. 22 act as an effective deterrent, even though in some cases these seem to be low).
  - Statistics are reviewed periodically by the National Statistics Commission (composed by the NSI, several public entities, universities, and representatives from the private sector and labor unions). The commission facilitates data sharing and coordination among statistics producing agencies, and also makes recommendations on processes and methodologies, but coordination is insufficient.
- Resources:
  - The price statistics program lacks adequate staff, facilities, and computing resources.
  - Insufficient staff precludes attaining a well-diversified outlet sample, a continuous analysis and implementation of methodological improvements, and appropriate quality adjustments to price statistics.
  - Computers and compilation software are outdated.
  - In the event of no access to the NSI’s facilities, compilation of the CPI would not be possible.
  - Office space and equipment are not adequate (more space, a cooling system, and telephone lines are required).
  - Resources to update regularly the business register and to conduct and validate surveys for national accounts statistics are also insufficient.
- Relevance:
  - The NSI has periodic meetings with the CBCH to identify emerging CPI requirements, but there is no structured consultation process for the PPI.
  - Public knowledge on the PPI is scant and index credibility is low.
  - The NSI seeks to implement international standards that should better attend users’ needs.
- Other quality management:
  - The NSI is quality conscious. Staff has been trained on quality management systems.
  - The NSI has not implemented an externally recognized process to focus on quality, but it is planning to implement an ISO 9001:2000 quality management system on the CPI by December 2007 and on the PPI by December 2008.
- Professionalism:
  - Law 17374 of 1970 recognizes the independence of the NSI to determine the methodology, release dates, design of surveys, and use of definitions. Data gathering is objective and impartial.
  - The NSI Director is appointed by the President of the Republic from a group proposed by the Civil Service Council. Deputy directors are elected by the director from a group proposed by the aforementioned Council.
  - Short tenure in both cases (three years plus one possible renewal) and discretion for removal are not supportive of professional independence.
  - Hirings and promotions are based on relevant aptitude, but there is a significant legal rigidity to terminate contracts with nonperforming staff members.
- Transparency:
  - The terms and conditions under which the statistics are produced and disseminated are available to the public.
  - Government officials outside the NSI have no access to the data before public release.
  - Vulnerability: staff from the Corporate Image Department (which does not belong to the Price Statistics Department) has access to CPI and PPI results prior to their release.
  - Major methodological changes are usually announced to the public in advance.
- Ethical standards:
  - The Administrative Statute Law, Law 18834 of 1989, details the public servants obligations and prohibitions, and forbids specific activities to avoid conflict of interest. Penalties are specified in cases of noncompliance.

### Budget Directorate of the Ministry of Finance (DIPRES) — Key findings
- Legal and institutional environment:
  - The Annual Budget Laws (Decree-Law 1263 of 1975 and Decree with Legal Power 106 of 1960) set a general framework for budgetary planning, execution, and reporting applicable to central government units, and provide the DIPRES of the MOF with the legal authority to compile and disseminate GFS.
  - The Law of Fiscal Responsibility (Law 20128 of 2006) empowers the DIPRES to collect data on transactions regulated by the Reserved Copper Law (Law 13196 of 1976), both for stocks and flows.
  - The Law on Municipalities (Decree with Legal Power 1/18695 of 2006) and the Law on Municipal Resources (Decree 2385 of 1996) set the responsibility of local governments regarding the preparation of annual financial statements, and reporting to the Office of the Comptroller General (OCG).
  - There is room for improvement in data sharing and coordination among the various units involved in the compilation process.
  - GFS are compiled using mostly publicly available data.
- Resources:
  - At present, the number of staff devoted to produce GFS is barely adequate.
  - In the medium term, additional resources may be necessary to complete the migration to the GFSM 2001 methodology.
  - Most staff is well trained in the GFSM 2001.
  - Physical resources are adequate.
- Relevance:
  - The DIPRES monitors the relevance of GFS through informal contacts with other departments and data users through e-mail inquiries, personal contacts, and press conferences.
  - There are no user advisory groups.
- Other quality management:
  - Quality is a main concern for GFS production.
  - Beginning in 2007, the DIPRES will be audited using the ISO 9001:2000 international quality standard.
- Professionalism:
  - The Administrative Statute (Law 18834 of 1989) and the Law of Administrative Probity (Law 19653 of 1999) assist in ensuring impartiality in the work of civil servants.
  - Professionalism is actively promoted and supported within the organization through external and on-the-job staff training.
  - Statisticians are free from political influence in choosing data sources and methods for compiling GFS, and changes to statistical processes are not influenced by political considerations.
  - The DIPRES is entitled to comment and respond to misused or misinterpreted fiscal statistics.
- Transparency:
  - Laws and regulations governing the compilation and dissemination of GFS are freely available to the public, with the exception of the Reserved Copper Law.
  - No government officials outside the DIPRES have access to the data before they are released to the public, and the data are made available simultaneously to all interested parties.
- Ethical standards:
  - Aside from the Administrative Statute and the Law of Administrative Probity, several codes of conduct governing the behavior of the central government and local government employees ensure that data compilation and dissemination are performed according to strictly professional considerations.

### Central Bank of Chile (CBCH) — Key findings
- Legal and institutional environment:
  - Provisions for producing statistics are included in the Law 18840 of 1989 (Basic Constitutional Act of the Central Bank of Chile).
  - Article 53 states that “the Bank shall timely compile and publish the main macroeconomic statistics, including those of a monetary and foreign exchange character, balance of payments and national accounts ...”
  - The National Statistics Compilation Plan assigns the CBCH the task of compiling and disseminating NA, monetary, and BOP statistics.
  - Article 53 establishes mandatory data reporting by public entities to the CBCH.
  - There are no legal precepts empowering the CBCH to require private institutions to respond solely to statistical data requests, but there is broad cooperation of the private sector with the CBCH.
  - Article 40 of the CBCH’s Basic Consitutional Act empowers it to demand from the private sector information on foreign exchange transactions.
  - Confidentiality of the reported data is guaranteed by Article 66 and the CBCH’s By-Laws.
  - The CBCH has special agreements with various institutions for data sharing, in particular with the NSI, the SBFI, the National Custom Service (NCS), the Internal Tax Service (ITS), and the National Tourism Service (SERNATUR).
- Resources:
  - Human resources dedicated to compile and disseminate macroeconomic statistics are adequate in number and professional qualifications.
  - Staff attend courses and training programs offered by international organizations, graduate programs abroad, and internships in recognized statistical agencies. Compensation levels are competitive.
  - Computing resources, physical infrastructure, and funding are adequate.
  - NA uses a data model and a database application to compile statistics. BOP is working on such an application.
  - A data warehouse application would make the database application more effective.
  - Measures are in place to monitor the cost-effectiveness of the statistical programs and to detect needs for improvements.
- Relevance:
  - The CBCH does not have formal mechanisms to consult data users about their information needs.
  - Users can submit questions and feedback on series and the database to the CBCH’s website.
  - The CBCH participates in statistical meetings and seminars organized by international and regional organizations.
- Other quality management:
  - The CBCH has a long tradition of producing high-quality macroeconomic statistics and strives to maintain high standards of quality using modern technology.
  - Among the objectives of its 2006–2008 strategic plan is to contribute to an understanding of economic and financial affairs by preparing and broadcasting relevant reports, studies, and statistics. Results of studies produced by the CBCH are disseminated to the general public.
- Professionalism:
  - By Constitution, the CBCH is an autonomous institution of a technical nature with its own capital (Chapter XII), guaranteeing professional independence.
  - The CBCH’s management promotes professionalism with a continuous training program and by receiving technical assistance from other countries and international organizations.
  - The CBCH monitors media coverage of macroeconomic statistics and provides comments and clarifications through its Communications Management Office when statistics are misinterpreted or misused.
- Transparency:
  - The CBCH disseminates its statistics in a timely manner and provides advance release calendars, which are strictly followed.
  - No government officials outside the CBCH have access to the data before their release to the public.
  - Documentation is available on the terms and conditions and the methodology used for NA, BOP, and monetary statistics.
  - Data released to the public are clearly identified as products of the CBCH.
  - Major changes in the conceptual framework, source data, and statistical techniques of macroeconomic statistics are announced in advance.
- Ethical standards:
  - The CBCH’s Personnel By-Laws provide guidelines for CBCH employees.
  - The Ethics Code (March 2003) establishes clear standards describing procedures for potential conflict of interest situations (Section II, 1) and includes measures to prevent the misuse or misinterpretation of statistics (Section II, 2). Compliance is facilitated by a culture of solid ethical standards.

### Superintendency of Banks and Financial Institutions (SBFI) — Key findings
- Legal and institutional environment:
  - The General Banking Act (Decree with Force of Law-DFL 3 of 1997) defines the SBFI as an autonomous institution, with full legal capacity, in charge of the supervision of the Banco Estado, the banking enterprises, and the financial entities whose control is not otherwise entrusted by the law to a different institution (Articles 1 and 2).
  - Article 16 of DFL 3 of 1997 indicates that “... the Superintendent shall have the right to request from the institutions subject to his supervision, any information, document, or book that, in his opinion, may be necessary for purposes of control or statistics.”
  - Penalties for not complying with the instructions issued by the SBFI are foreseen in Article 19 of DFL 3 of 1997.
  - Confidentiality of the information provided to the SBFI is guaranteed by Article 7 of DFL 3 of 1997.
  - Several agreements of collaboration between the SBFI and the CBCH are in place, under the umbrella of Article 53 of Law 18840/89.
  - The SBFI forwards electronically to the CBCH daily and monthly data received from other depository corporations (ODCs), via dedicated encrypted lines.
- Resources:
  - Human, technical, and financial resources are generally adequate.
  - Salaries in the SBFI are regulated by the scale of public supervisory institutions and are higher than the rest of the public sector. However, they are lower than salaries paid by the private banking sector, inducing a higher than desirable personnel turnover.
- Relevance:
  - The SBFI’s main function is to collect data for supervisory purposes.
  - The SBFI has demonstrated willingness to accommodate the CBCH’s emerging data needs for the compilation of monetary statistics, requesting additional information from the ODCs when necessary.
- Other quality management:
  - The SBFI’s authorities are aware that continued efforts are needed to improve the quality of its statistical products.
- Professionalism:
  - The SBFI is a technical institution that has administrative and financial autonomy.
  - Technical criteria alone are applied by data compilers and analysts, and they are totally independent in their choice of data collection methods.
  - Processes and activities in the workplace promote a culture of professionalism.
  - Staff of the SBFI regularly attend meetings and participate in international courses offered by other supervisory agencies and central banks.
- Transparency:
  - The legal terms and conditions for collecting data for supervisory purposes are specified in the relevant laws, which are available on the SBFI’s website.
  - According to Article 14 of DFL 3 of 1997, the Minister of Finance may have access to information on entities supervised by the SBFI, but this provision refers basically to cases of financial insolvency and does not affect published monetary data.
  - On its website, the SBFI disseminates data on its supervised institutions, such as balance sheets, financial statements, interest rates, and composition of deposits and loans.
- Ethical standards:
  - Through internal staff bulletins, SBFI staff is made aware of its rights and obligations.
  - Special attention is paid to the confidentiality of the information received by the SBFI, with concrete instructions on the ways to protect it.

*Source: _cr07318 - 0. Prerequisites of quality (PDF chapter content as provided).*

### 2. Methodological soundness

### 2. Methodological soundness

### National Accounts (NA)
- Concepts and definitions
  - Compiled following the 1993 SNA.
- Scope
  - Regular compilation of annual and quarterly GDP by production and expenditure approaches at current and constant 2003 prices.
  - Annual supply-and-use tables, input-output matrix for the base year, and accounts by institutional sectors up to the financial account.
  - Indirect estimates on nonobserved activities derived by applying the commodity flow approach.
  - Some exceptions to 1993 SNA treatment: mineral exploration, agricultural work in progress, and own-account production in some activities.
  - Free zones and workers who work part of the year in other countries are included.
- Classification/sectorization
  - Broadly consistent with ISIC Rev. 3 and CPC; deviations include classification of copper derivatives in mining.
  - COFOG and COICOP are not used for government and household final consumption classifications; improvements under development.
- Basis for recording
  - Transactions recorded on an accrual basis, except for agricultural products.
  - Transactions between establishments of the same enterprise recorded on a gross basis.
- Source data and assessment
  - NSI business register partially updated; HBS conducted every 10 years (instead of five).
  - Income tax and VAT records used to fill gaps for agriculture, fishing, some services, and quarterly estimates.
  - Source data reasonably approximate NA requirements; timeliness could be improved.
  - Validation measures exist but NSI resource constraints limit timely validation; scientific sampling techniques not used for most surveys.
- Statistical techniques
  - Broadly appropriate with minor deviations.
  - GDP compilation mainly based on supply-and-use tables; changes in inventories are residuals.
  - Double-deflation method used to estimate around half of value added.
  - Only some explicit estimates on nonobserved activities for the benchmark year.
  - Base year updated to 2003.
  - Denton method used for benchmarking quarterly and annual figures.
  - Quarterly GDP data are seasonally adjusted.
- Consistency, revisions, and dissemination
  - Preliminary GDP data meet SDDS recommendations on periodicity and timeliness.
  - NA series consistent from 2003 to 2006; previous series consistent for 1996–2005.
  - Intertemporal and internal consistency verified in supply-and-use tables.
  - Net lending/borrowing by institutional sector derived independently and reconciled; differences between NA and GFS reconcilable.
  - BOP statistics and NA broadly consistent.
  - Preliminary data clearly identified; preannounced calendar of revisions.
  - Base year changed every seven to 10 years historically; plan to update every five years.
  - Documentation on analysis of regular and major revisions is prepared and made public.
- Accessibility and metadata
  - NA disseminated in clear manner but with insufficient detail, in hard copy and on CBCH’s website.
  - Advance release calendar gives a 12-month-ahead notice.
  - Simultaneous release to all users; customized tabulations provided upon request.
  - Detailed updated metadata available in special publications and on CBCH’s website; metadata on IMF’s DSBB need updating.
  - No specific contact person indicated; users can address queries via an e-mail address and Communications Management Office.
  - Catalog of CBCH publications available; users can subscribe to an e-system for notifications.

### Consumer Price Index (CPI)
- Concepts and definitions
  - Broad correspondence with CPI Manual, 2004.
  - CPI basket includes interest rates on consumption loans and mortgage payments; treatment inconsistent with best practices (combined weight about 3.6 percent).
- Scope
  - Measures price change for a fixed basket representative of households of Greater Santiago (about 40 percent of Chile’s population), excluding other cities and rural areas.
  - NSI conducting a new national HBS; plan to update CPI base with expenditure results by December 2008.
- Classification/sectorization
  - Domestic classification closely corresponds with COICOP but with departures (e.g., food includes restaurants and alcoholic beverages; education bundled with recreational services).
  - Classification structure: eight groups, 41 subgroups, 156 articles, 482 products, and more than 1,300 varieties.
  - Domestic classification of educational services splits product payments for the same service, producing potential downward bias; education services weight 6.2 percent.
- Basis for recording and source data
  - CPI weights based on consumption expenditure valued at purchaser’s prices.
  - Expenditure estimates for used cars follow an erroneous netting procedure yielding a relatively high index weight for this item (1.1 percent).
  - CPI basket and weight structure based on 1996–1997 HBS.
- Assessment and statistical techniques
  - Various validation procedures for source data; no established procedures to verify HBS response consistency with other sources.
  - Sound data compilation and appropriate methods for missing prices.
  - New products introduced in sample as they gain market share, but similar treatment for outlets is lacking.
  - Adjustments for quality differences are not made (except for computers, where the procedure requires revision), potentially biasing the index with new products.
- Intermediate validation and revisions
  - Intermediate results regularly assessed and validated.
  - Thresholds for investigating unusual movements are invariant to item characteristics, leading to under- or overvalidation.
  - NSI has not conducted long-term trend studies for HBS revision patterns; HBS levied on 10-year cycles, leading to infrequent weight revisions.
- Periodicity, consistency, accessibility
  - CPI published monthly within five working days after month end, meeting SDDS requirements.
  - Statistics consistent within dataset and over time; methodological notes identify main breaks.
  - Unusual changes in price trends not explained in press release.
  - CPI final upon first release.
  - Punctual publication on preannounced schedule, but presentation hinders interpretation and meaningful comparisons; press release lacks appropriate analysis.
  - NSI provides comprehensive CPI metadata on its website.
  - Adequate assistance to users.

### Producer Price Index (PPI)
- Concepts and scope
  - Output weights determined according to internationally accepted standards.
  - PPI covers six categories: agriculture, livestock and forestry; mining; fishing; manufacturing; electricity, gas and water; and construction. It does not cover services.
  - Monthly surveys collect prices for about 4,000 specific goods in 738 establishments.
- Classification/sectorization and basis for recording
  - Built on classification systems consistent with international standards; ISIC Rev. 3 used for activities, CPC ver. 1.0 for products.
  - Exception: classification of copper extraction and copper manufacturing departs from international classification to conform to CBCH national accounts.
  - Procedures to value flows and timing rules in accordance with 1993 SNA, but survey design does not allow registry of copper prices on accrual.
- Source data and assessment
  - Comprehensive data collection programs; respondents aware of reporting deadlines.
  - Sample excludes firms with less than 10 employees, enterprises within free zones, and household unincorporated enterprises.
  - Export and domestic prices considered but not weighted according to sales.
  - NSI Methodology Department routinely reviews censuses and surveys; prices examined for consistency with other sources.
- Statistical techniques and validation
  - Sound imputation techniques for missing prices in agriculture and livestock and forestry; in other categories missing prices are carried forward (practice recommended to discontinue).
  - Adjustments for quality differences are not made; copper price recollection procedure increases quality-induced volatility because quotes are contract-based with quality adjustments.
  - Intermediate results regularly assessed and validated; investigative thresholds invariant to item characteristics.
  - Sources of errors are not routinely investigated.
- Periodicity, consistency, accessibility
  - PPI published monthly within five working days after month end, meeting SDDS requirements.
  - Statistics consistent within dataset and over time; unusual changes not explained in press release.
  - PPI final upon first release.
  - Statistics published on schedule but presentation hinders interpretation and comparisons; press release lacks appropriate analysis.
  - An indicator excluding copper effects on volatility is not published; detailed product price series not available on NSI’s website.
  - Summary of PPI methodology published on NSI’s website; full version not available.
  - Adequate assistance to users.

### Government Finance Statistics (GFS)
- Concepts and coverage
  - Concepts and definitions broadly consistent with GFSM 2001; DIPRES documented official migration plan to GFSM 2001 since 2003.
  - Institutional and transactional coverage incomplete: universities and ancillary units serving local governments not covered.
  - Central government debt excludes accrued interest on debt instruments, liabilities related to the Reserved Copper Law, and the stock of recognition bonds (supplementary data disseminated separately for latter).
- Classification and basis for recording
  - Sectorization broadly in line with GFSM 2001 but significant deviations in national practices (copper revenues, subsidies and grants, capital transfers, and financing transactions do not conform).
  - Tax revenues recorded on a cash basis.
  - Interest of indexed debt instruments does not include changes in principal due to index fluctuations.
- Source data and assessment
  - Source data complete and timely for GFS compilation.
  - SIGFE provides comprehensive accounting for budgetary central government; stock data recorded at face value.
  - Extrabudgetary central government data collected by DIPRES; local governments provide data via OCG.
  - Debt GFS compiled using Treasury accounting records.
  - Built-in software cross checks monitor data accuracy.
- Compilation, validation, and revisions
  - Data compilation mostly computerized.
  - Process for validating data not well established.
  - GFS data are final when first published; major revisions linked to methodological changes are documented and disseminated.
- Periodicity, consistency, accessibility
  - GFS meet SDDS standards; Chile takes a flexibility option for timeliness of central government operations.
  - Consistent time series available from 1990; monthly data available from 2003.
  - GFS not internally consistent due to differences in source data and valuation methods used to compile debt flows and stocks.
  - Discrepancies between GFS and NA explained by coverage, sectorization, and classification differences; residence criteria differences explain inconsistency with BOP.
  - GFS disseminated on DIPRES website in clear manner; publications released on regular preannounced schedule, most in Excel-readable format.
  - Non-routine statistics available upon request.
  - Metadata on concepts, scope, classifications, recording basis, data sources, and methodologies widely documented by DIPRES and available in publications, on website, and on DSBB.
  - Contact persons and details publicized in DSBB metadata; catalogs of publications are not available to statistics users.

### Monetary Statistics
- Concepts and classification
  - Monetary statistics in line with MFSM methodology.
  - Coverage of depository corporations (DC) survey complete, including CBCH, commercial banks, state-owned Banco Estado, credit unions, and mutual funds.
  - Classification of financial instruments follows MFSM; securities issued by CBCH and liabilities of mutual funds are not sectorized.
- Basis for recording and valuation
  - Produced on an accrual basis, consistent with MFSM.
  - Foreign currency accounts converted into national currency at end of month using CBCH’s “observed” exchange rate (average of mid-point exchange rate of interbank market).
  - Three nonmaterial departures from MFSM:
    1. Financial instruments valued at lower of amortized cost or market prices, rather than market prices.
    2. Interest not accrued on nonperforming loans, recorded in memorandum accounts.
    3. Gold holdings recorded using average of daily prices of last three months, rather than closing day quotation.
- Source data and assessment
  - DC survey based on consolidated balance sheets of CBCH and ODCs.
  - ODCs report daily preliminary data to CBCH and monthly balance sheets plus sectorized data to SBFI, which forwards to CBCH.
  - Data transmitted electronically minimizing risk of error; validated by SBFI including interbank positions.
- Techniques, validation, and revisions
  - Monetary data based on balance sheets; GEMF uses additional data to sectorize ODCs’ data; standardized macros run in processing.
  - GEMF analyzes daily data and investigates large unexplained fluctuations but does not validate aggregated interbank positions, holdings of securities by ODCs, or monetary data against fiscal data.
  - GEMF routinely cross-checks net foreign assets with IIP but does not reconcile net credit to government with fiscal data.
  - Inconsistencies detected in aggregated interbank positions due to lack of sectorization of mutual funds’ liabilities.
  - CBCH has produced documented revisions (e.g., changes in monetary aggregate definitions) published in CBCH Estudios Económicos Estadísticos.
  - Monetary data published provisional for two months, then substituted by final data; provisional data and revisions clearly identified.
- Periodicity, consistency, accessibility, metadata, and assistance
  - Monetary statistics disseminated on the 7th and 23rd day after the reference month, in full compliance with SDDS requirements.
  - Data disseminated in consistent historical series; CBCH and ODC survey data internally consistent.
  - Data posted on CBCH’s website, Boletín Mensual, and Síntesis Monetaria y Financiera; CBCH publishes and complies with an advance release calendar.
  - Monetary metadata posted on IMF’s DSBB since February 1997; CBCH website links to DSBB country page.
  - Contact information for monetary statistics available on DSBB; users’ requests and comments may be channeled through CBCH’s website.
  - Catalog with CBCH publications, documents, frequency, and prices is available.

*Source: _cr07318 - 2. Methodological soundness*

### 2. Methodological soundness

### 2. Methodological soundness

### Concepts and definitions
- Balance of payments (BOP) broadly follow the BPM5.
- Resident units are defined in conformity with the BPM5.
- In 2002, the CBCH completed the implementation of the BPM5 and migrated balance of payments data to the BPM5 from 1996 onwards.
- Scope: BOP is broadly consistent with international standards. No significant amounts of unrecorded activities are deemed to occur.

### Classification and sectorization
- Classification/sectorization is broadly in line with the BPM5 guidelines.
- Reverse investment on FDI are an exception.

### Basis for recording
- For the most part, transactions/stock data are valued at market prices.
- In general, transactions are recorded on an accrual basis in accordance with the BPM5.
- Exceptions where accounting is done on a different basis:
  - Loan liabilities of public sector and private sector are recorded on a cash basis.
  - Portfolio and other investment income are generally recorded on a cash basis for liabilities and on an accrual basis for assets.
  - Unrequited transfers, such as grants, are recorded when foreign exchange is converted into the national currency (pesos).
  - External trade is recorded according to customs documents.
- Entries of income are registered on accrual basis with the exceptions of external debt.
- Gross and net transactions are recorded in accordance with the BPM5.

### Source data
- Main sources are broadly adequate to compile balance of payments and include:
  - the CBCH’s Summary of International Exchange Regulations,
  - customs data,
  - balance sheets,
  - administrative data,
  - surveys and estimates.
- These data are supplemented or validated with information from enterprises, foreign agencies, and international financial institutions.

### Assessment of source data and statistical techniques
- Source data are routinely assessed, including for surveys used by the Balance of Payments and External Debt Department (DBPDE).
- Robust and efficient techniques are used by the DBPDE to adjust data derived from primary sources to improve quality, coverage, and timelines.
- Estimates are used to adjust data in case of nonresponse or missing data.
- Computerized edit checks are used where database management procedures exist to identify coding and other errors.
- Statistical discrepancies are assessed and validated with available datasets.
- The DBPDE routinely carries out revisions of balance of payments data to improve its statistics.
- The CBCH publishes on its website studies on specific related topics and explanations of the causes of revisions.

### Periodicity, timelines, and consistency
- BOP and international investment position (IIP) statistics meet SDDS requirements.
- Partial data publication frequency:
  - weekly (external trade),
  - semi-monthly (reserve assets),
  - monthly (financial account flows).
- Balance of payments data are broadly consistent internally and over time (annual data since 1996 and quarterly data since 2002).
- BOP statistics are consistent with the IIP and monetary statistics.
- A full reconciliation with national accounts statistics was attempted.
- External debt statistics are not consistent with BOP because they value the portfolio liabilities on a nominal basis.

### Revision policy and practice
- Data are provisional when first released.
- Revisions follow the announced calendar on the Boletín Mensual.
- Quarterly data of the calendar year are revised each time a new quarter is released.
- Annual revisions are carried out in March of the following two years.
- The public is informed of major changes in methodology on the CBCH’s publications.

### Data and metadata accessibility; user assistance
- Data are disseminated in accordance with the BPM5 standard components.
- Quarterly release is made under the analytical presentation; the annual version uses a detailed presentation.
- Data are made available to all users at the same time on the CBCH’s website and released on a preannounced schedule.
- Concepts, methodology, and data sources for balance of payments compilation are available on the CBCH’s website, in the Balance of Payments of Chile publication (annual), and on the IMF’s DSBB.
- Assistance to users is adequate. Publications on the CBCH’s website provide an institutional contact point; for specific balance of payments statistics, users’ contact points are provided on the DSBB.

### Cross-cutting and institutional recommendations (selected high priority)
- Ensure that the NSI has adequate financial resources, staff, facilities, and training, and take further steps to increase retention of qualified staff.
- Strengthen independence of the NSI by lengthening tenure of the NSI Director and Deputy Directors from three to six years, staggering appointments, maintaining the reappointment option, and limiting discretion for removal.
- Enhance coordination among statistical agencies through closer contact and regular meetings of technical staff to discuss specific data requirements, reduce respondent burden, improve timeliness of source data, and avoid duplication.
- Reinforce the legal framework of the CBCH to ensure its legal mandate to request statistical data from the nonfinancial private sector.
- Initiate regular consultations with public and private sector users, including fostering users’ groups to improve the usefulness of statistics.
- Intensify efforts to improve intersectoral data consistency and reconciliation by addressing differences in coverage, classification, and sectorization among datasets.

### Balance of Payments Statistics — recommendations
High priority
- Coordinate with the ITS to reduce the time to access tributary data records to improve the quality of quarterly BOP estimations.
- DIPRES should consider the CBCH’s data requirements for statistics purposes (BOP, IIP, external debt, and NA) to distinguish the residence of the payments under the framework of the Reserved Copper Law, in the context of the Art. 19 of the Law of Fiscal Responsibility.
- Design and implement a data warehouse for external statistics (BOP, IIP, and external debt) to facilitate DBPDE’s data handling; current storage is in a large number of electronic sheets with complex and unstable links.
- Record under Reserve Assets the net creditor balance on the Latin American Integration Association reciprocal credits agreements (A accounts minus B accounts).
- Record transactions related to the external debt on an accrual basis.

Other recommendations
- Follow the guidelines of the BPM5 in applying the directional criteria to reverse investments transactions of resident foreign direct investment enterprises using available information at the CBCH.
- Reclassify trade credit flows of enterprises included in the foreign direct investment directory as foreign direct investment liabilities flows in the case of imports with customs’ information recorded under the “collect mode.”

*Source: _cr07318 - 2. Methodological soundness*

### APPENDIX I

### APPENDIX I

### SDDS Coverage, Periodicity, and Timeliness — Key dataset metadata (Table 4 excerpts)
- National accounts
  - SDDS: Coverage (meets SDDS requirement): Yes; Periodicity: Q; Timeliness: Q
  - Chile: Periodicity: Q; Timeliness: Q; Timeliness details: 7W; 11W for Q4
- Production index / Global production index
  - SDDS: Coverage: Yes; Periodicity: M; Timeliness: M
  - Chile: Periodicity: M; Timeliness: M; Timeliness details: 6W (1M encouraged); 4W; 35D
- Forward-looking indicators (encouraged data category: economic expectations, consumer confidence (CC), index of perception of the economy (IPE))
  - SDDS: Periodicity: M or Q; Timeliness: M or Q
  - Chile: Periodicity: M, Q; Timeliness: CC 2Q; IPE 36D
- Employment
  - SDDS: Coverage: Yes; Periodicity: Q; Timeliness: Q
  - Chile: Periodicity: Q; Timeliness: Q; Timeliness details: 4W
- Unemployment
  - SDDS: Yes; Q; Q
  - Chile: Q; Q; 4W
- Wages/earnings
  - SDDS: Yes; Q; Q
  - Chile: Q; Q; 5W
- Consumer price index
  - SDDS: Yes; Periodicity: M; Timeliness: M
  - Chile: M; M; 5D
- Producer price index
  - SDDS: Yes; M; M
  - Chile: M; M; 5D
- General government operations (Fiscal)
  - SDDS: Yes; Periodicity: A; Timeliness: A
  - Chile: Periodicity: A; Timeliness: 2Q; Timeliness details: 6M
- Central government operations
  - SDDS: Yes; Periodicity: M; Timeliness: M
  - Chile: M; M; 1M — Chile is availing itself of the flexibility option.
- Central government debt
  - SDDS: Yes; Q; Q
  - Chile: Q; Q; 3M
- Analytical accounts of banking sector
  - SDDS: Yes; Periodicity: M; Timeliness: M
  - Chile: M; M; 3W
- Analytical accounts of the central bank
  - SDDS: Yes; Periodicity: M (1W recommended); Timeliness: 2W
  - Chile: Periodicity: M; Timeliness: 2W (1W encouraged); 1W
- Interest rates; Exchange rates; Stock market: share price index
  - SDDS: Coverage: Yes; Periodicity: D; Timeliness: D
  - Chile: D; D; D (where applicable)
- Balance of payments
  - SDDS: Yes; Q; Q
  - Chile: Q; Q; 7W — Fourth quarter, 83 days; other quarters, 53 days.
- Official reserve assets / Reserves template
  - SDDS: Coverage: Yes; Periodicity: M (W recommended); Timeliness: M
  - Chile: Periodicity: M; Timeliness: W; Chile timeliness examples: M 1W; 23 days
- Merchandise trade (exports: f.o.b; imports: f.o.b)
  - SDDS: Yes; Periodicity: M; Timeliness: W
  - Chile: Periodicity: M; Timeliness: M; Timeliness details: 8W (4–6W encouraged); 1W; 2W
- International Investment position
  - SDDS: Yes; Periodicity: A (Q recommended); Timeliness: 6M
  - Chile: Periodicity: A; Timeliness: 3Q (1Q encouraged); 2Q
- External debt
  - SDDS: Yes; Periodicity: Q; Timeliness: M
  - Chile: Q; M; Q; 5W
- Addendum: Population
  - SDDS: Yes; Periodicity: A; Timeliness: A
  - Chile: A; A; 18M

Note: Periodicity and timeliness codes from source: (D) daily; (W) weekly or with a lag of no more than one week; (M) monthly or with a lag of no more than one month; (Q) quarterly or with a lag of no more than one quarter; (A) annually; (...) not applicable. Italics in the source indicate encouraged categories.

### Cross-cutting recommendations and authorities' responses (high priority and selected actions)
- Ensure INE has adequate resources, staff, facilities, and training; increase retention of qualified staff.
  - INE: considered an increase in the budget for next four years to create a CPI study department and retain qualified staff participating in new CPI 2008; request included in INE’s budget proposal for next year.
- Strengthen independence of INE by lengthening tenure of INE Director and deputies from three to six years, staggering appointments, maintaining reappointment option, limiting discretion for removal.
  - Authorities: matter beyond INE’s authority; requires Law of Civil Service changes.
- Enhance coordination among statistical agencies via regular technical meetings, reduce respondent burden, improve timeliness and avoid duplication.
  - Authorities: coordination emphasized between INE and BCCH; to be enhanced for 2008 benchmark national accounts compilation.
- Reinforce BCCH legal framework to request data from nonfinancial private sector.
  - BCCH: Article 53 of BCCH founding law authorizes requesting information from public entities; no amendment planned; BCCH uses INE sector surveys intensively and currently obtains satisfactory response for its surveys despite lack of explicit private-sector legal mandate.
- Initiate regular consultations with public and private users; foster user groups.
  - Authorities: INE creating a Committee of Users of INE Statistics; Central Bank to include formal user consultations beginning in 2008.
- Intensify intersectoral data consistency and reconciliation (coverage, classification, sectorization).
  - Authorities: creation of Committee INE-BCCH to revise classifiers/nomenclatures; first reconciliation results expected by end of the year.
- Disseminate contact persons responsible for each dataset.
  - Authorities: INE will include technical record, e-mail addresses, telephone numbers in publications; BCCH will include webpage and e-mail addresses in Monthly Bulletin and Yearbooks; DIPRES will disseminate contact names in new statistical publications.

### Central Bank of Chile — Responses and work program highlights
- National accounts (high priority)
  - Update business register regularly by requesting annual sales and employment by enterprise to the ITS; conduct simplified economic census every five years.
    - INE leads National Directory of Establishments; actions to collect information from Internal Revenue Service (SII).
  - Clarify INE responsibility for collecting source data for national accounts in INE’s law revision.
    - INE law revision (bill) intended to strengthen independence, statistical secrecy, and professional staff.
  - Apply ISIC Rev. 3 and CPC classifications; publish complementary tables for copper-related activities and main export products.
    - 2008 benchmark compilation: examine statistical separation of copper activity between mining production and refined goods (manufacturing); methodological research planned.
  - Increase level of detail of disseminated data at three digits of ISIC or at least two digits.
    - 2008 benchmark compilation will increase entries of products and activities.
  - Reconstruct historical series when methodology/source changes occur; review linked series for 1996-2005.
    - 2008 work program includes review of statistical retropolation methods.
- Monetary statistics
  - Properly sectorize securities issued by BCCH and liabilities of mutual funds.
    - BCCH developing methodology to identify holders of its documents; consultations with SVS regarding mutual fund liabilities identification.
  - Perform systematic validations of reported monetary data within and across datasets.
    - Multiple validation stages exist; BCCH will explore additional information to permit further validations.
  - Compile other financial corporations survey (pension funds, insurance corporations, other intermediaries).
    - BCCH will examine feasibility to make better compilations for significant financial sectors.
- Balance of payments statistics (high priority)
  - Coordinate with ITS to reduce time to access tributary data to improve quarterly BOP estimations; DIPRES to consider BCCH data requirements for residence distinctions under Reserved Copper Law and Art. 19 of Law of Fiscal Responsibility.
    - BCCH will request SII to evaluate shortening timeliness of reception of information; will resume request to DIPRES for complementary information on Reserved Copper Law operations.
  - Design and implement a data warehouse for external statistics (BOP, IIP, external debt).
    - Bidding process underway to hire external consulting to develop procedures and valuation of desired model for handling storage; results will serve as basis for an information system.
  - Record under Reserve Assets the net creditor balance on Latin American Integration Association reciprocal credits agreements (A minus B accounts).
    - BCCH: matter is numerically insignificant (less than 1 percent of reserves stock); timing of adoption to be assessed considering BPM6 and IFRS effects and publication costs.
  - Record external debt transactions on an accrual basis.
    - Project of 2006 produced results from 2004 onwards; BCCH intends to incorporate accrual recording into BOP and external debt statistics in March 2008, conditional on Statistics Committee approval, with revised figures for 2005 and 2006.
- Other BOP recommendations
  - Apply BPM5 directional criteria to reverse investments of resident FDI enterprises using BCCH information — included in 2007-2008 work plan.
  - Reclassify trade credit flows of enterprises in FDI directory as FDI liabilities for imports recorded under "collect mode."
    - Integrated Short-Term Debt Data Base project launched to calculate commercial debt at micro level.

### National Statistics Institute (INE) — Price statistics responses and actions
- Cross-cutting IT and metadata
  - Provide adequate protection for computer resources; emergency back-up system currently inside building; INE evaluating alternatives to move back-up outside building and will buy a new operating system to be in a different location.
  - Determine procedures to adjust prices for quality changes.
    - Staff compiling, studying, and evaluating best international practices to generate customized procedures.
  - Improve presentation and content of press release.
    - Continuous improvement process; expected maturity by end of 2007.
  - Develop a more user-friendly webpage.
    - INE will carry out a survey of users and start a process to develop a user-friendly webpage; creating a webpage for students with easy charts and historical series.
- Price statistics — management quality (other recommendations)
  - Establish ISO 9001:2000 management quality system on planned schedule.
    - Price Department started initial stages in June 2007; planned schedule to finish end of 2008 with external consulting support.
  - Redefine thresholds for unusual price movements by idiosyncratic characteristics of each item.
    - INE reviewing CPI processes; will redefine thresholds for each product in basket.
  - Upgrade facilities: additional office space, two telephone lines for PPI field agents, cooling system.
    - Efforts underway to make improvements.
- Consumer Price Index (CPI) — high priority actions
  - Update CPI weights and items in the basket (current structure based on 1996-1997 expenditure patterns).
    - Since October 2006 INE updating CPI weights and items; process to finish by November 2007. HBS compiled with national coverage. INE requesting resources to carry out HBS every five years.
  - Expand geographical scope beyond Great Santiago.
    - Planned expansion to all regions in Chile; expansion will have national representativeness but not regional in CPI 2008. Budget to hire qualified regional surveyors at end of 2007; requesting resources to make this permanent.
  - Revise methodology of financial expenditures and mortgage payments per CPI Manual, 2004.
    - INE revising methodology according to CPI manual and international experiences.
  - Reclassify educational services at product level to merge payment types toward same service to avoid downward bias.
    - To be considered in CPI 2008; product-level weights limit reclassification.
  - Amend methodology to derive used cars weight adopting international standards.
    - INE accepts recommendation; will consider in CPI 2008; as first stage, used cars will be left out of basket because used car flows to Chile are negligible.
  - Analyze, refresh, and expand outlet sample.
    - R&D Department tasked to analyze, refresh, and expand outlet sample; new proposal under study.
  - Review HBS consistency with alternative sources to verify CPI weights (alcoholic beverages, tobacco).
    - INE working to obtain alternative data sources to verify CPI weights and incorporate adjustments.
  - Publish X1 price index (BCCH core inflation definition) on INE’s website.
    - X1 is published by BCCH; INE will discuss with BCCH to have X1 published by INE in the future.
- Producer Price Index (PPI) — high priority actions
  - Conduct survey among users to identify reasons for low credibility; implement measures.
    - NSI preparing a survey and will hold meetings among data users; findings to be incorporated into improvement process.
  - Review methodology for copper price records to address quality and volatility.
    - Methodology under revision; experts hired on the topic.
  - Revise methodology to handle missing prices in mining; manufacturing; electricity, gas, and water; construction.
    - New PPI planned for 2009; methodology to be revised in all aspects.
  - Publish a PPI version excluding copper and emphasize it in press release.
    - Pilot tests conducted to produce index without copper to avoid volatility.
  - Publish detailed PPI historical time series on INE website.
    - Action taken; historical series available on INE webpage.
- PPI — other recommendations
  - Consider expanding PPI coverage to include services.
    - INE evaluating inclusion of services.
  - Publish detailed metadata on INE webpage.
    - Price Department evaluating human and financial resources; planned completion by end of 2007.

### Budget Office (DIPRES) — Government finance statistics (GFS) actions and priorities
- High priority recommendations to implement
  - Follow GFSM 2001 guidelines in classifying revenue, expenses, transactions on nonfinancial and financial assets and liabilities, for stocks and flows.
  - Improve institutional and data coverage of general government:
    - Include universities and ancillary units serving local governments.
    - Record capitalized interest on fiscal promissory notes with BCCH (subordinated debt) as interest expense in statement of central government operations.
    - Include as liabilities of central government interest accrued and not paid for all debt instruments, and the stock of recognition bonds.
  - Reconcile debt stocks and flows periodically to improve internal horizontal consistency of GFS.
- Other recommendations and actions
  - Follow GFSM 2001 in valuing stocks of assets and liabilities at market basis or nearest equivalent.
  - Record interest on indexed debt instruments per GFSM 2001 principles.
    - DIPRES aiming to establish cooperation agreement with BCCH to formalize data-sharing, enhance coordination in data collection, and address remaining data lacks.
  - Institutional measures
    - DIPRES considering implementation of a user group to advise DIPRES on unmet user needs.
    - New statistical publications will disseminate contact persons responsible for each dataset and include a catalog of statistical publications.
    - Committee established to assess mission issues and propose implementation schedule to Budget Director; coordination with SIGFE on automatic consolidated output is part of committee tasks.
  - DIPRES note on indexed debt interest treatment
    - DIPRES maintains that GFSM 2001 principles advise uneven treatment for operations with same practical effects.

*Source: APPENDIX I, Chile — Report on the Observance of Standards and Codes (ROSC)—Data Module; Responses by the Authorities; Detailed Assessments Using the Data Quality Assessment Framework (DQAF); August 30, 2007.*

### 1. National Accounts Source Data ...................................................................................26

### 1. National Accounts Source Data

### I. Prerequisites of quality — Overview
- Detailed assessment uses the Data Quality Assessment Framework (DQAF, July 2003) structure.
- Compiled and disseminated by the Central Bank of Chile (CBCH); National Accounts Department (NAD) of the Research and Statistical Information Management Office (GIIE) is the compiling unit.
- NAD organization: six working groups handling compilation and reconciliation of production, goods and services accounts, and institutional accounts.
  - Five groups compile and reconcile production and goods and services accounts (including sectoral production accounts for benchmark, annual, quarterly series, and IMACEC).
  - One group compiles and reconciles annual institutional accounts and regional GDP.
- Benchmark compilations impose significant temporary increases in staff time (professional staff time increases by practically 50 percent for a two-year period versus normal schedule).

### 0.1 Legal and institutional environment — Key findings
- Legal mandate and responsibilities:
  - Constitutional Law of the CBCH (Law No. 18840 published in the official gazette on October 10, 1989): Article 53 — “the Bank shall compile and publish, in a timely manner, the country’s principal macroeconomic statistics, including monetary, foreign exchange, balance of payments statistics, and national accounts.”
  - CBCH Board of Directors establishes coverage, periodicity, and timeliness; agreements are published in the official gazette.
  - Law of the National Statistics Institute (NSI) (No. 17.374 of October 15, 1970): Article 1º establishes NSI as institution in charge of official statistics and censuses; Article 2º lists functions including collection, technical elaboration, analysis and publication of official statistics, coordination of statistics compiled by fiscal and semi-fiscal entities, conducting official censuses, conducting surveys to update indicator bases (in particular cost of living), preparing registers of statistical information sources, and submitting annually the National Plan of Statistical Collection for presidential approval.
- Practical arrangements:
  - NSI is responsible for surveys and censuses supplying source data for NA; responsibility not always explicitly specified in NSI law beyond cost of living.
  - CBCH finances collection of source data or specific breakdowns that are not part of NSI customary production due to NSI budgetary and human resource constraints.
  - NSI is preparing a draft of a new NSI law to update legal framework.

### 0.1.2 Data sharing and coordination among data-producing agencies — Findings
- Major source data producers:
  - NSI: sectoral economic surveys on production, household surveys, prices, compensation of employees, employment.
  - CBCH: conducts complementary surveys and finances noncustomary NSI products.
  - Administrative records: National Customs Service (NCS) (customs records) and Internal Tax Service (ITS) (tax records).
- Formal agreements and contracts:
  - Framework agreement for annual, quarterly, and monthly follow-up compilations (1994), updated annually.
  - Framework agreement for benchmark compilations (2003) defining specific NSI studies.
  - Service delivery contracts for other information (e.g., purchasing power parities study, 2004).
  - Conventions and contracts establish quality and timeliness terms for NSI products.
  - Specific conventions cited: Conventions of September 14, 1994; No. 0869 of March 01, 2003; No. 1059 of October 19, 2004; No. 1407 of November 13, 2006; No. 1304 of March 20, 2007.
- Institutional coordination bodies:
  - Permanent Committee (quarterly meetings) and Coordination Committee (monthly meetings) for monitoring agreements.
  - Interministerial Committee: Director of the NSI, President of the CBCH, and Minister of Finance.
  - National Statistics Commission (CNE) of the National Statistics System (NSS) encourages interagency cooperation; meets every two months.
  - NSS includes NSI, CBCH, National Office of Planning, CORFO, DIPRES (MOF), academia, producers’ associations, and labor unions (labor union representative has not attended most CNE meetings).
- NAD contributions to interagency coordination:
  - Developed initial business register/directory of enterprises (2002) and transferred to NSI for management (2006).
  - Established system of classifiers of products, activities, and agents; updates centralized by Inter-Agency Nomenclature Committee.
- Gaps and issues:
  - Some duplication of effort in collecting source data (notably mining and employment).
  - Improvements in NSI timeliness have been significant, but data validation and basic-data quality improvements are gradual due to NSI resource constraints.

- Recommendations (from source):
  - Request data on annual sales and employment by enterprise from the ITS to update the business register regularly and gross-up sample data, in line with “Statistical Secret” precepts in NSI law and confidentiality in CBCH law.
  - Conduct simplified economic census (precensus) every five years to update sampling frames.
  - Enhance coordination among statistical agencies through closer contacts and regular meetings of technical staff to discuss specific data requirements at a detailed level, reduce respondent burden, improve timeliness of source data, use collected data in all their potential, and avoid duplication of effort.

### 0.1.3 Confidentiality and data access — Findings
- Legal and institutional protections:
  - Article 66 of the Law of the CBCH guarantees confidentiality of data reported by individuals and entities.
  - CBCH Manual of Ethics Code and By-laws prohibit officials from divulging confidential information or unpublished information; employment contracts establish confidentiality obligations and specify termination for violations.
  - NSI law Articles 29 and 30: strict “Statistical Secret”; violation is criminal under Article 247 of the Criminal Code; individual statistical data shall not be published without authorization.
  - NSI questionnaires include statutory obligation and confidentiality statements referencing Law No. 17,374.
  - CBCH surveys enclose confidentiality letters; surveys mainly filled via Internet with individual passwords known only to respondents.
- Technical and organizational security measures at CBCH:
  - Hierarchical system of privileges for protected access to information sources and processes.
  - Department of Operations and Technology implemented security measures: passwords/access lists, intranet access with authorized user name and password, different user profiles via Windows XP, external firewall, antispam module, antivirus.
  - Encrypted mail sharing system for confidential/highly sensitive information.
  - Manual of Policies on Informatics Security (March 2003) establishes policies on “clean desks,” Internet/e-mail use, norms for treatment of confidential/highly sensitive information, and mechanical destruction of hard-copy information.
  - Questionnaires archived in CBCH storeroom and destroyed after five years upon request.
- Access to administrative records:
  - Customs records: expeditious access without confidentiality restrictions.
  - Tax records (ITS): available on an anonymous and aggregate basis based on classifiers established by NAD due to confidentiality of individual records; data sharing agreed through technical meetings and recorded in minutes.
- Assessment: Confidentiality safeguarded through legal provisions and technical controls.

### 0.1.4 Statistical reporting mandate and response rates — Findings
- Legal authority:
  - CBCH authorized under Article 53 of its law to obtain information from public institutions for compiling the NA.
  - No legal means empowering CBCH to require private institutions to respond to data requests for compiling the NA; nevertheless, broad cooperation exists from private agents.
- Observed response performance for CBCH surveys (2006):
  - Response Rate to CBCH’s Surveys for 2006
    - Communication: No. of surveys 121; Response rate (in percent) 77; Weighted response (in percent) 94
    - Construction: No. of surveys 366; Response rate (in percent) 77; Weighted response (in percent) 81
    - Health: No. of surveys 140; Response rate (in percent) 52; Weighted response (in percent) 97
    - Electricity: No. of surveys 103; Response rate (in percent) 83; Weighted response (in percent) 85
    - Mining: No. of surveys 53; Response rate (in percent) 89; Weighted response (in percent) 98
    - Fishing: No. of surveys 71; Response rate (in percent) 56; Weighted response (in percent) 56
    - Health (second line): No. of surveys 90; Response rate (in percent) 46; Weighted response (in percent) 82
    - Services: No. of surveys 3; Response rate (in percent) 67; Weighted response (in percent) 52
    - Transport: No. of surveys 485; Response rate (in percent) 45; Weighted response (in percent) 89
    - Total: No. of surveys 1,432; Response rate (in percent) 62; Weighted response (in percent) 89
    - Note: Weighted response rates are weighted by output of each enterprise of the sample.
- Issues:
  - Lower response rates in fishing and some services surveys.
  - Some surveys too detailed; data collected not always used to full potential due to timeliness or data quality/confidence issues.
- Recommendation (from source):
  - Strengthen the legal framework of the CBCH to ensure legal mandate to request data from the nonfinancial private sector.
  - Adapt the level of detail of some surveys to the needs of benchmark and annual follow-up compilations to reduce respondent burden and improve timeliness and usability of collected data.

### National accounts compilation practices — Key operational notes
- Use of administrative records: quality and coverage are good; appropriate filters developed for use in NA compilations.
- NAD handles multiple periodicities: benchmark, annual, quarterly, and monthly (IMACEC). Quarter/benchmark processing requires reconciling many sources and time series of differing periodicities, justifying organizational arrangement with multiple groups.
- CBCH finances NSI products not part of customary NSI production and specific breakdowns.

*Source: _cr07318 - 1. National Accounts Source Data (Extracted content provided).*

### 0.2       Resources

### 0.2       Resources

### 0.2.1 Staff, facilities, computing resources, and financing are commensurate with statistical programs
- The NAD of the CBCH has a staff of 30 professionals, 12 one third of whom have been compiling the NA for 10 or more years, while the rest have been doing so for no more than five years on average.
- Staff composition and training:
  - Staff are professionals, primarily economists, with a lesser number of statisticians.
  - Approximately one-third of staff have post-graduate degrees at either the masters or doctoral level.
  - The NAD develops orientation courses for new professionals and training workshops on statistical and accounting methods, sends staff to short-term seminars (ECLAC, CEMLA, FIIAPP-Madrid), medium-term courses (IMF), and emphasizes on-the-job training.
  - Recent training priorities include visits to BEA, StatCan, U.K. ONS, UNSD, Eurostat; participation in OECD seminars; long-term internships with INE (Spain) and StatCan; and masters’ degrees at Columbia University and the University of Nottingham.
  - Footnote: Plus five new hires contracted, initially, for a fixed term.
- Compensation and human resources management:
  - Compensation levels and structure are competitive with the market, particularly if nonmonetary benefits are included.
  - All staff were reclassified to make salary mobility more flexible.
  - The salary scale is currently being adapted to goals and to provide performance bonuses for achieving such goals.
- Computing and technical resources:
  - Desktop computers are at least a Pentium 4HT, with networked Windows XP Operating System.
  - Hardware includes color printers and photocopiers, black and white printers and copiers for heavy-duty work, portable computers, and overhead projectors.
  - Software: Microsoft Office (Excel, Word, PowerPoint, Access), and analytical tools eViews, MatLab, SAS. Plan to incorporate FAME/SunGard as the database for time series.
  - Surveys’ systems platform: Web three layers, Client Browser, and Component Services, as well as client/server with an application developed in Visual Basic 6.0 and databases developed in SQLSERVER 7.0.
  - NA compilation platform: developed in Access databases and Excel worksheets.
  - All staff connected to the Internet; active Intranet for compilation and communications processes.
  - Adequate protection and emergency back-up systems in place; back-ups are daily updated.
- Financing:
  - Funding for the compilation of NA is provided under the budgetary process of the CBCH.
  - Short- and medium-term work programs are taken into account in the budgetary process.
  - Requests for funding for special projects (changing the benchmark year of the accounts, expanding coverage, conducting methodological studies) are included as part of the normal budgetary process.

### 0.2.2 Measures to ensure efficient use of resources are implemented
- Strategic planning and resource allocation:
  - The CBCH has a three-year (2006–2008) strategic plan that serves to increase the quality and timeliness of compiled statistical series and to improve efficiency in resource allocation.
  - The three-year plan is broken down into annual work plans administered by each organizational unit in coordination with the Administration and Managers Committee of the CBCH.
  - The Administration and Managers Committee meets every quarter to review progress and discuss allocation of financial and human resources.
  - Each unit formally reviews progress on a monthly basis according with their work plans.
- Staff evaluation:
  - Staff are evaluated annually based on integration, contributions, and achievement of preestablished goals.
  - The evaluation method is being revised to align with new salary mobility and performance bonus practices.
- Database and process improvements:
  - A NA database has been developed based on a specific data model tying transaction-related data to attributes (agent, type of operation, type of object) to improve consistency between classification concepts and reconciliation of benchmark and annual follow-up compilations.
  - Extensions under development:
    - Sectoral compilation: data model for various industries to improve sectoral compilation and preliminary reconciliation.
    - Processes model: documentation on processes and each compilation frequency may result in a processes model complementing the data model.
    - Consideration to move from Access to other platforms (SQL Server) to increase efficiency, scale, and execution speed.
- Benchmark compilations:
  - Benchmark compilation has been carried out every seven to 10 years and consistently reviewed by outside experts for methodology and compilation systems.
  - The benchmark compilation is the most costly compilation and compares favorably in cost-benefit terms with previous exercises in Chile and the region.
  - The current strategic plan includes updating the benchmark year to 2008 (every five years).
- Budgetary formulation:
  - The CBCH has an elaborate process for budgetary formulation and monitoring of budgetary execution to facilitate resource allocation and achievement of results within the strategic planning framework.

### 0.3 Relevance

#### 0.3.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored
- User feedback mechanisms:
  - The CBCH’s website provides two e-mail addresses for users to submit feedback on the statistical database and other suggestions.
  - No other specific measures (surveys, informative bulletins, seminars) have been adopted to ensure statistics meet users’ needs.
  - The NAD has no periodic and direct process for consulting with substantive users (NA users, academia, media, private sector) to evaluate usefulness or identify emerging requirements.
- Responsiveness:
  - High degree of sensitivity and response capacity to requests from users and CBCH’s own research and management units.
  - Examples: IMACEC (1985) developed to meet monetary planning policy needs; quarterly accounts (2005) developed to meet inflation-targeting monetary policy requirements.
- Participation in external forums:
  - NAD actively participates in statistical meetings and seminars organized by regional organizations (ECLAC and Andean Community (CAN)) and has begun joining organizations outside the region (OECD).
  - Footnotes: Chile rejoined the CAN early this year. Chile’s incorporation as an extra-regional member of the OECD will be examined next month.
- Recommendation:
  - Initiate regular consultations with public and private sector users to identify new and emerging data requirements.

### 0.4 Other quality management

#### 0.4.1 Processes are in place to focus on quality
- Strategic objectives (2006–2008):
  - Objectives: “prepare and disseminate to the country on a timely basis national, regional, and sectoral statistics on areas within the purview of the Central Bank of Chile and on the conduct of its policies, maintaining high standards of quality and using the best technology available.”
- High-impact NAD initiatives:
  - Developing and implementing the 2008 benchmark compilation (CR2008).
  - Improving the quality of statistics through specific projects in the NA areas.
- Commitment:
  - GIIE and NAD committed to achieving objectives and establishing specific goals for all NAD staff; units elaborate working plans and training programs.

#### 0.4.2 Processes are in place to monitor the quality of the statistical program
- Monitoring procedures:
  - Procedure 1: Structure initiatives (improving quality through specific NA projects) based on IMF focus on gaps between local situation and best practices in sources, methods, and results; IMF recommendations from this report will aid in adjusting and prioritizing initiative components.
  - Procedure 2: Maintain history of different product versions using BEA review of results approach for quarterly and annual follow-up compilations; methodology added for systematic comparison between the 2003 benchmark compilation and the 2003 follow-up compilation (base year 1996); applied for 2004 and 2005 by comparing rebasing results according to 1996 and 2003 benchmarks.
  - Procedure 3: Monitor procedural quality following the Eurostat approach based on documentation of processes; may result in a processes model complementing the data model, allowing comparison between periods and versions based on sources (response rate) and methods (double deflation versus single indicator) at most disaggregated level.
- Benefits:
  - Procedural quality evaluation would allow assigning a reliability level to each record used in compilation and contribute to objective measurement of data quality for preliminary reconciliation purposes.
- Dissemination and revision cycle:
  - The dissemination and revision cycle of the NA constitutes an adequate procedure for monitoring quality in terms of factors explaining discrepancies between preliminary and revised estimates and allows steps to reduce discrepancies when feasible.

#### 0.4.3 Processes are in place to deal with quality considerations in planning the statistical program
- Program development:
  - NA program development incorporates quality improvements identified in follow-up and periodic revisions and user feedback on quality standards and new data requirements.
  - Decisions weight various quality dimensions considering available resources.
  - Approach: incorporate best level (accuracy/reliability) measurements in less frequent measurements (annual or benchmark) and measure changes from best level through more frequent measurements (quarterly and monthly) to optimize cost/benefit of the national information system for NA purposes.

### 1. Assurances of integrity

#### 1.1 Professionalism

##### 1.1.1 Statistics are produced on an impartial basis
- Legal basis and institutional independence:
  - Professional independence of the CBCH is established in Article 1º of its Constitutional Law.
  - Article 1 states the CBCH is an autonomous institution that is technical in nature, of unlimited duration, has juridical personality, and its own capital.
  - Impartiality in NA production is a CBCH feature of the highest order and a prerequisite for monetary policy effectiveness.
- De facto impartiality:
  - Although no specific statutes prevent external influence on NA content and dissemination, in practice there is no external or internal influence impacting NA results.
- Governance and tenure:
  - Article 8 of the CBCH’s law: Board members’ tenure is 10 years, appointments may be renewed; replacements staggered every two years.
  - Board composed of five members appointed by the President of Chile with Senate approval.
  - Governor appointed by the President from among Board members for five years or remainder of board term.
- Internal safeguards:
  - Standards establish professional behavior, courtesy toward respondents, integrity, impartiality in hiring, execution of official duties, and avoidance of third-party influence.
  - Since October 2005, GIIE reports to General Management instead of Research Division to strengthen internal independence.
  - A CBCH Statistics Committee ensures formal relations between GIIE and the rest of CBCH.
  - Recruitment and promotion based on professional aptitude and expertise; vacancies posted on CBCH’s website.
  - Contract hiring considers professional profiles for specific jobs.
- Professional development and research:
  - Staff encouraged to attend seminars, pursue training, and conduct research projects (case studies for supply-and-use table consistency; evaluation of compilation in real terms using chain indices).
  - Staff prepare analytical texts on NA topics for CBCH publications and journals.

##### 1.1.2 Choices of sources and statistical techniques as well as decisions about dissemination are informed solely by statistical considerations
- Selection criteria:
  - Selection of source data and statistical techniques is based solely on statistical considerations, measurement objectives, and data requirements.
  - Cost considerations may sometimes determine selection of sources and methods.
  - Decisions to disseminate data are based on technical, timeliness, and cost considerations.
- Independence:
  - NA are compiled in accordance with professional criteria and with complete independence.

##### 1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics
- Communication and correction:
  - CBCH responds to erroneous interpretations through press releases and other means.
  - Data presented to the media at press conferences held at release times; comments on GDP trends included in presentations and the CBCH’s Monthly Bulletin distributed to the press to identify factors behind unusual figures, movements, and revisions.
  - The President and the manager of the GIIE are entitled to respond to public criticism or misuse of statistics; few instances have required such response.

#### 1.2 Transparency

##### 1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public
- Public information:
  - CBCH’s website includes its law, which informs about CBCH standards of confidentiality and professionalism; law available in print in Spanish or English from the CBCH’s Publications Department.
  - The CBCH By-Law and the Manual of Ethics Code posted on the CBCH’s intranet.
  - CBCH website contains a directory of key personnel and two e-mail addresses for requests and contact.
  - CBCH publications include contact information; CBCH library provides public information.

##### 1.2.2 Internal governmental access to statistics prior to their release is publicly identified
- Pre-release access:
  - NA are released simultaneously to all users; no person or institution outside the CBCH has access prior to publication.
  - Exception: the Minister of Finance attends some Board meetings and receives the IMACEC—under embargo—24 hours in advance of general release.
  - This access is informed in the metadata for NA of the SDDS posted on the IMF’s DSBB; the CBCH website links to the SDDS website.

##### 1.2.3 Products of statistical agencies/units are clearly identified as such
- Identification:
  - All CBCH publications include the bank’s name and logo and follow a standard format (titles, colors, typography).
  - Joint publications identify all institutions involved.
  - CBCH requests source acknowledgement when third parties reproduce NA records.

##### 1.2.4 Advanced notice is given of major changes in methodology, source data, and statistical techniques
- Advance notice practices:
  - Users informed in advance primarily through articles in the Monetary Policy Report (IPOM), briefings, or news releases of major changes.
  - Example: new series of NA at 2003 prices published on March 23, 2007 was reported in October 2006 and internalized by users in advance.
  - It has been reported that changes can be expected in NA sources and methods with publication of the 2008 Benchmark Compilation (CR2008) to be published in 2011.

*Source: _cr07318 - 0.2       Resources*

### 1.3       Ethical       standards

### 1.3       Ethical       standards

### Ethical standards and staff behavior
- A Manual of Ethics Code (March, 2003) establishes clear standards describing procedures for potential conflict of interest situations (Section II, 1).
- The Manual includes clear ethical standards on measures to be adopted to prevent the misuse or misinterpretation of statistics (Section II, 2).
- Implementation is facilitated by a solid ethical standards culture that discourages political interference.
- The manual is available on the CBCH’s intranet and staff are made aware of the manual in orientation courses when they join the organization.
- The CBCH law, its By-Laws, and the Manual of Ethics Code include guidelines on staff behavior regarding discretion over affairs reserved by their character and observance of decorous conduct and respect with peers and the public.
- The manual states incompatibilities and conflicts of interest that are prohibited, for example:
  - Receiving from parties with business relationships with the CBCH any invitation or gift that can be related to expedite paperwork or transactions.
  - Any other circumstance that could induce a special treatment or inappropriate authorization.
- Obligations of staff include:
  - Being efficient in the performance of their duties.
  - Updating their knowledge regularly.
  - Using the bank’s goods properly.
- Sanctions for non-compliance (presented in the CBCH’s By-Laws) include:
  - Verbal reprehension.
  - Written reprehension.
  - Dismissal of the employee.

### Methodological soundness — overall framework
- The general framework used for compiling the national accounts is guided by the recommendations of the System of National Accounts 1993 (1993 SNA).
- Implementation of the 1993 SNA has been gradual in the 1986, 1996, and 2003 benchmark compilations.
- Major deviations from concepts and definitions established in the 1993 SNA involve:
  - Agricultural work-in-progress.
  - Investment in in-house development of software.
  - Mining exploration.
  - Measurement of rentals of owner-occupied housing.
- These issues are kept under review and are on the agenda for the 2008 benchmark compilation.

### Scope — compilation frequencies and minimum/recommended requirements
- Follow-up compilations are produced based on the benchmark compilation annually, quarterly, and monthly.
- 1993 SNA accounts and tables determined by ISWGNA as minimum and recommended requirements are compiled as follows:

  Annual follow-up compilation:
  - Annual value added and GDP at current and constant prices by economic activity
  - Annual value added components at current prices by activity
  - Annual expenditures of GDP at current and constant prices
  - Annual rest-of-the-world accounts (up to net lending)
  - Sequence of accounts for the total economy (up to net lending) with an annual frequency
  - Annual supply-and-use tables (recommended requirement)

  Quarterly follow-up compilation (recommended requirements):
  - Quarterly value added by activity and GDP at current and constant prices
  - Quarterly expenditures of GDP at current and constant prices

  Monthly follow-up compilation:
  - Volume indicator of economic activity, a synthetic benchmark indicator measured of production volume by economic activity

### Scope — production and assets boundaries (deviations and inclusions)
- Production boundary is generally consistent with the 1993 SNA, with exceptions attributed more to lack of information than conceptual deviations:
  - Own-account production of manufactured goods for own-final consumption is excluded; own-account production of agricultural goods for own-final consumption is included.
  - Research and development activities on own account are included as costs in intermediate consumption of the principal activity and not as individual products.
  - Mining exploration is not included in investment, although the costs thereof are measured.
  - Production of entertainment, literary, or artistic originals is not included (item is very insignificant).
  - Production of computer software for own use is not included (not a significant item).
  - No estimates are made of illegal output sold to willing buyers and of shuttle trade/contraband.
- Delimitation of constituent units of the economy follows the 1993 SNA:
  - Free zones operated by offshore enterprises under customs control and workers who work part of the year in other countries are included as part of the economy.
- Assets boundary is broadly consistent with the 1993 SNA:
  - Tangible assets include national defense related assets that could be used for civilian purposes.
  - Valuables and historical monuments are not included.
  - Regarding agricultural, livestock, and forestry work-in-progress, only forestry and livestock work-in-progress is included.
- Intangible assets treatment:
  - Mineral exploration is measured as intermediate consumption and not as investment.
  - Computer systems, software, and databases are included as assets only when purchased from third parties and not when companies develop them in-house.
  - Entertainment, literary, or artistic originals are not included.
  - Patented entities, financial leases, and other transferable contracts are not included.

- Recommendation:
  - Incorporate agricultural work-in-progress and mineral exploration in the production and assets boundary, given the importance of mining and agriculture activities in the Chilean economy.

### Classification / sectorization
- Classification and sectorization used for compiling the national accounts are broadly consistent with recommended international systems, with some deviations:
  - The 1993 SNA is used to classify institutional units, transactions, and other flows.
  - ISIC Rev. 3 is used to classify principal economic activities (industries) of establishments and enterprises, with exceptions (e.g., inclusion of copper-manufactured products in mining).
  - Correlative codes are used instead of the international standard codification.
  - A local adaptation of the Central Product Classification (CPC), called the Single Product Classifier, is used to classify products; correlative codes are used instead of the international standard codification.
- Deviations in classifications by function or purpose:
  - COICOP is not used for classifying household consumption expenditures, but it is being implemented.
  - COFOG is not applied to classify functions of government, but it is being implemented.
- Given the relational nature of the NAD databases, implementation of different classifiers is handled by correspondence tables linking one classifier to another, allowing compilation and reconciliation to continue and results to be reclassified in terms of recommended international systems.
- Recommendation:
  - Apply the standard ISIC Rev. 3 and CPC for classifying activities and products, respectively, to facilitate international comparisons, and publish complementary tables for domestic analysis of copper-related activities and main export products.

### Basis for recording
- Market prices are used to value flows and stocks in accordance with 1993 SNA recommendations.
  - Market output is valued at basic prices.
  - Output for own use is valued at equivalent market prices.
  - Taxes on products are part of the valuation of production at producer prices and of intermediate and final consumer goods (in Chile this involves taxes on tobacco products and petroleum refining products).
  - Value-added taxes are included in the valuation of intermediate consumption in exempt activities but the deductible portion of the value added tax on other activities is excluded.
  - The deductible portion of value added taxes is excluded from the valuation of final uses.
  - Corrections are made when transfer prices are detected to reflect market prices (applies in particular to outputs from vertically integrated industries).
  - Total imports and exports are valued on an f.o.b. basis.
  - Transactions in foreign currency are converted using the mid-point exchange rate prevailing in the market at the moment they take place or by using monthly average exchange rates.
- The above apply equally to the benchmark compilation and the annual follow-up compilation.

- Recording basis:
  - Transactions and flows are recorded on an accrual basis.
  - Government revenues, which are on a cash basis, are allocated to the corresponding period to convert them to accrual basis.
  - Government expenditures are recorded on an accrual basis.
  - Major exception: agricultural work-in-progress — agriculture output is recorded at the time of the harvest instead of at the time it occurs. This deviation is relevant for the compilation of quarterly GDP.

- Grossing/netting procedures:
  - Grossing/netting procedures are consistent with the 1993 SNA.
  - As of the 2003 benchmark compilation, all transactions between establishments within the same enterprise are recorded on a gross basis.

*Source: _cr07318 - 1.3       Ethical       standards*

### 3.1       Source       data

### 3.1       Source       data

### 3.1.1    Source data collection, coverage, and institutional strategy
- Source data for the National Accounts (NA) in Chile are developed on a decentralized basis (censuses, surveys, administrative records) independently of the NAD; development of source data for NA purposes is uneven.
- Greater availability of source data for measuring GDP:
  - Production approach > Expenditure approach > Income approach (relative reliability).
- Within production approach, source data stronger for traditional activities (agriculture, mining, manufacturing) and weaker for some service industries.
- Periodicity of source data:
  - Production and other indicators: monthly product-level statistics available.
  - Expenditure data: partially included on a quarterly basis.
  - Income data: incorporated on an annual basis.
  - Availability and reliability "largely follows international standards."

- NAD strategy to address weaknesses:
  - Intense use of the supply-and-use table (SUT) to integrate source data.
  - Relate sources across time series with different periodicity; statistics collection program is medium-term and coincides with Chile’s NA compilation cycle: benchmark compilation → annual, quarterly, monthly follow-up compilations.

- Major institutional management initiatives:
  - Establish cooperation and technical assistance with sources outside the NAD: design forms, validation processes, attribution of annual statistics from enterprises/establishments, household surveys prices (NSI), public finances (DIPRES, MOF), and structure classifications/filters for administrative records (ITS and NCS).
  - Where necessary, carry out directly or via third parties:
    - Periodic data collection activities for annual, quarterly, and monthly follow-up compilations.
    - Special studies primarily for the benchmark compilation.
    - Improving quality of source data.

- IMACEC sources:
  - Use direct indicators of production volume when available (monthly statistics on manufacturing and mining production) requiring reclassification and concept adjustments.
  - Indirect indicators are used to construct direct indicators when needed (e.g., agricultural, livestock, fishing).
  - Building permits and square meters built used for construction estimates.
  - Export indicators used where product is primarily sold abroad.

- Quarterly accounts sources:
  - Exports and imports by product available monthly but used quarterly.
  - Product sources available quarterly: register of investment projects (projects > five million dollars) and quarterly survey of manufacturing and trade (used mainly for cross-checking).
  - Good quarterly data on mining, electricity, gas, and water; some transportation services; government information timely and of good quality; health indicators limited to public sector.
  - Lack of information on many other services; these are estimated using VAT records.
  - Expenditure side: monthly index of retail sales of consumption goods compiled; quarterly survey on inventories; imports classified by economic use.

- Annual accounts sources:
  - More product detail (e.g., annual agriculture survey on crops and yields).
  - Construction data include CBCH survey on main engineering projects.

- Data collection effort for benchmark compilation:
  - CBCH processed 3,282 financial statements and 24,916 establishment/enterprise surveys for benchmark compilation efforts.
  - Benchmark compilation used extensive NSI surveys, special surveys, and case studies (e.g., Agriculture, Fruit-Farming, and Livestock Survey; Survey of Small Enterprises; construction study).

- Household Budget Surveys (HBS):
  - Conducted every 10 years (not every five). Last HBS was August 1996–July 1997 (Greater Santiago) used in 1996 benchmark; a new HBS is currently being conducted.

- Summary assessment in text:
  - (i) Product sources satisfactory with limited exceptions: agricultural products, fishing, administrative records on construction permits, internet services, and HBS.
  - (ii) Industry sources satisfactory and some good, except real estate and housing services. Annual surveys on manufacturing, domestic trade, services, and restaurants and tourist accommodation are mixed: not used to full potential in annual compilation due to timeliness/quality, but satisfactory for benchmark compilation where NAD can devote resources. Grossing-up factors use tax records for business and personal services and some manufacturing.

### 3.1.2    Source data conceptual consistency and coverage
- Source data "reasonably approximate" NA definitions, scope, classifications, valuation, and time of recording required.
- Recording bases:
  - Annual data from establishments and nonfinancial enterprises: accrual basis; disaggregated according to private sector business accounting.
  - Financial enterprises: accrual basis per their chart of accounts.
  - Government: expenditures on an accrual basis; revenues on a cash basis but allocated to the corresponding period.
- Reclassification to 1993 SNA is feasible.

- Coverage assessment (Table 4 summary):
  - Coverage categories: L: less than 50 percent, M: 50-79 percent, H: 80% or more.
  - Coverage of total economic activities in annual surveys is satisfactory except for domestic trade, restaurant/hotel, transport, business, real estate, housing, and other personal services.
  - Income tax records used to improve coverage in these activities.
  - Activities still limited after grossing-up: agriculture, fishing, real estate and housing services.

- Table 4 (coverage shorthand preserved):
  - Examples from Table 4:
    - Agriculture: M (product method) for Annual Survey and Tax Records.
    - Mining: H H H.
    - Manufacturing: M M H M H.
    - Real estate and housing: L L L.
    - Public administration: H H H.

### 3.1.3    Timeliness of source data
- Overall timeliness: "no major problems" for financial and insurance companies, prices, foreign trade, balance of payments, general government, and CBCH surveys.
- Timeliness issues: manufacturing, domestic trade, some services could be improved; response to CBCH surveys of nonfinancial private enterprises is voluntary, causing some untimeliness; NSI resources insufficient for regular establishment visits.
- Recent improvement: monthly compilation release date moved forward from 53 to 35 days (two years prior to report).
  - When delays occur, CBCH uses related indicators, extrapolates trends, or uses partial information.

### Data collection quality, surveys, and institutional capacity
- NSI surveys:
  - National manufacturing survey collects detailed accounting-based information: operational income/expenditures, employment and compensation, production and sales of major products (volume and value), inputs quantity/value, inventories, investment by type, taxes, depreciation.
  - Surveys of trade and services mirror manufacturing survey structure but do not collect outputs and inputs due to service nature; special questionnaire for restaurant/tourist accommodation enterprises.
  - NSI annual surveys do not cover all service activities; NAD conducts complementary surveys.
  - NSI sample-frame procedure: census of major establishments, random samples for medium and small establishments; sample frames not regularly updated.
  - NAD developed a directory of enterprises from tax records, commercial guides, and surveys; enterprises classified by principal activity (ISIC Rev. 3). CBCH transferred responsibility for managing/updating directory to NSI in 2006; directory was updated with 2003-2004 data but only partially updated since then due to budget constraints and confidentiality of some income tax declaration data.

- Limitations in NSI capacity:
  - Timeliness of NSI survey results improved from 12 months to seven months, but validation still limited by NSI resource constraints.
  - NSI resource constraints affect staff training and retention of qualified staff; CBCH makes partial use of NSI surveys.
  - Some surveys are very detailed, increasing respondent burden and reducing timeliness.

- Recommendation (explicit in source):
  - Ensure that the NSI has adequate financial resources, staff, and training, and take further steps to increase retention of qualified staff.

### Source data valuation and deflation practices
- Price statistics for constant-price compilation are "partially adequate."
- For sectors with production measured by administrative records (electricity, communication, transportation), constant values estimated by deflation method.
- Imports and exports mainly deflated by unit value indices.
- Exports: also have price data for domestic production from establishment-based surveys.
- Constant price estimates of engineering works obtained by deflation using indices of construction costs.

### Benchmark compilation role and data intensiveness
- Benchmark compilation updates production matrix and absorption matrix completely and seeks best level for macroeconomic aggregates in the SUT; it is the primary occasion for full utilization of detailed product and industry information.
- Additional benchmark sources noted: 2003 Trade Margins Survey, Fishing Survey, 2002 Population and Housing Census, and greater use of annual statistics and administrative records.
- CBCH allocates more resources for data validation/processing during benchmark compilation; NSI annual surveys are fully utilized only at benchmark.
- Significant processing for benchmark: 3,282 financial statements; 24,916 establishment/enterprise surveys.
- Agricultural census: last full agricultural census supporting 1996 benchmark; a new agricultural census underway.

- Use of HBS in benchmarks: 1996–97 HBS used in 1996 benchmark; a new HBS currently being conducted.

*Source: _cr07318 - 3.1       Source       data*

### 3.2 Assessment of source data

### 3.2 Assessment of source data

### Coverage, data quality assessment, and validation of source data
- Source data types assessed: censuses, sample surveys, administrative records (including VAT, income tax records, yearbooks, statistical reports).
- Routine assessment practices:
  - NAD examines incoming information and carries out informal consultations with respondents to verify high value transactions.
  - Time devoted to evaluating annual data is greater than for monthly and quarterly data.
  - Source data are routinely validated for temporal consistency.
  - Information about imputations made to source data is available; household surveys are audited to verify accuracy.
  - The NAD keeps internal archives on encountered problems (e.g., misclassification, measurement).
- Limitations and gaps:
  - Information is not available about sampling and nonsampling errors for all surveys; sample errors, nonsampling errors and response errors are not available for most surveys.
  - Most establishments are not selected from probabilistic samples due to the lack of an up-to-date business register; therefore, coverage cannot be accurately assessed.
  - Validation of NSI surveys is affected by resource constraints, hindering the quality of source data for manufacturing, domestic trade, and some services.
- Administrative-records practices:
  - Statistical procedures detect atypical values and anomalous differences in VAT records.
  - Exactitude of foreign trade statistics, volumes, and unit value indices is evaluated regularly.
  - Improvements in scope have significantly increased quality of gross fixed capital formation measurement.

### Preannual reconciliation, benchmark analysis, and handling of atypical values
- Preannual reconciliation phase:
  - Source data are analyzed to correct deficiencies, control consistency with previous declarations, and ensure consistency with other related source data.
  - Effects of changes to questionnaires are evaluated when new surveys/questionnaires are incorporated.
- Treatment of atypical values:
  - Atypical values are not replaced or modified unless clearly warranted.
  - Compilation procedures minimize processing errors (coding, editing, tabulation).
  - Procedures for imputation and adjustment for nonresponse are soundly based.

### Benchmark compilation (2003) — methodology, scope, and balancing
- Historical context:
  - The 2003 benchmark compilation is the fifth in Chile: previous efforts in 1962, 1977, 1986, and 1996.
  - Compilation cycle reduction: from 10 (1986–1996) to seven years (1996–2003).
  - 2003 benchmark replaced estimates carried out for 2003 under the 1996 benchmark; annual, quarterly, and monthly estimates were reviewed and 2004–2006 accounts revised in nominal terms and rebased in real terms.
- System and structure:
  - Main compilation system based on a SUT disaggregated by activities and products.
  - Aggregated information organized in seven main tables to ensure consistency.
  - For the benchmark, data for 73 products and 73 industries were compiled.
- Balancing process and implications:
  - Supply-use balance achieved (supply equals use) via simultaneous balancing of products and industries, modifying the least reliable source components.
  - Commodity flow method used where independent expenditure estimates are not available; inventories mainly obtained as residuals and show erratic figures, trends, and implicit price indices.
  - Recommendation: "Improve estimates on changes in inventories by using the available data by product and analyzing their price movements."

### Coverage of nonobserved activities and small businesses
- Benchmark compilation includes some explicit estimates on small businesses but does not include explicit estimates on nonobserved activities (informal, hidden, unrecorded), which are only considered implicitly in SUT reconciliation.
- Potential implications of incomplete coverage: biased growth rates, misleading information on economic structure and GDP level, biased international comparability, distortions in internal consistency of the NA.
- Note: nonobserved activities, although not quantified, "might not be significant in Chile."
- Recommendation: "Improve estimates of nonobserved activities by using available employment and income data from household/employment surveys."

### Production approach procedures and specific estimations
- Detail and levels:
  - Output, intermediate consumption, and value added estimates compiled at three-digit ISIC (73 activities and 73 products), with source data processed for more than 200 activities at four-digit ISIC and some 1000 products.
- Use of ratios and valuations:
  - In the total estimate of value added, the use of fixed ratios is around 40 percent; these ratios were updated in 2003 and are less than five years old.
  - Owner-occupied dwellings: output valued as estimated rentals tenants would pay for similar accommodation.
- Work-in-progress and inventories:
  - Work-in-progress estimated only for standing timber (forestry), stocks of fish, and livestock reared for food.
  - Work-in-progress of growing crops and large construction projects are not calculated.
  - Partial inventories data used; output and intermediate consumption adjusted for holding gains/losses where applicable.
- Capital consumption and construction:
  - Perpetual inventory method partially used as conceptual basis for consumption of fixed capital.
  - Construction estimated using permits without adjustment instead of survey data.
- Tax and margin volume measures:
  - Volume measures of taxes/subsidies on products estimated by applying base-year-tax rates to transaction volumes.
  - Trade margins’ output volumes extrapolated using base-year-trade margins and volume extrapolators of sales.
- Deflation and volume measures:
  - Double indicator method (double deflation) used to estimate 45 percent of GDP at constant prices.
  - Single indicator method used for the remaining 55 percent.
  - GDP volume change is not calculated using annual chain indices although annual data by industry are available.
  - CBCH plans to change the base year on a five-year basis.
  - Recommendation: "Consider the use of chain indices for calculating volume measures."
    - Implementation needs: review annual survey data on production and costs by industry to collect main output and input costs; allocate more resources to the NSI for collection, processing, and validation of source data; PPI weights would need to be annually chained.

### Expenditure approach procedures and price measures
- Expenditure estimation practices:
  - Household final consumption expenditure, gross-fixed capital formation, and changes in inventories are not independently estimated; calculated using commodity flow method.
  - Last HBS conducted during 1996-1997 in the Greater Santiago area; therefore, household final consumption expenditure for the new series is based on an outdated HBS. A new HBS is currently being conducted.
  - Household final consumption compiled at product level but not reclassified at least at one-digit COICOP; government final consumption not compiled at least at one-digit COFOG (work underway).
  - Gross-fixed capital formation compiled by activities and by type of assets only for benchmark/base year (2003). Changes in inventories compiled by products, not by activities and type of inventories.
  - Government final expenditure excludes incidental sales. Residents’ expenses abroad included in household final consumption and imports; nonresidents’ expenses in the economy excluded from household final consumption and included in exports. Expenditures on stores of wealth (jewelry, works of art) not included in valuables.
- Price indices:
  - Wholesale Price Index (WPI), 1992=100, CPI, and other dedicated price indices used to deflate GDP components at least at the one-digit level.
  - Producer Price Index (PPI), 2003=100, recently updated but "is not widely used even though the WPI is extremely outdated."
  - Recommendation: "Widen the use of the Producer Price Index (PPI) when compiling estimates at constant prices."
  - Household consumption implicit deflator consistent with CPI.
  - Government final consumption deflated by cost components of output for final use; government value added extrapolated by employment.
  - Changes in inventories: values at current and constant prices estimated in SUT reconciliation; deflators are implicit price indices.

### Quarterly and monthly compilation techniques
- Quarterly:
  - Quarterly production and expenditure approaches combined using a partial SUT that considers supply-use identities at product level.
  - No independent estimate of GDP by expenditure; commodity flow method assigns use based on nature of good and SUT identities used for residual determination.
  - For quarterly compilation, complete information for production or use matrices is not available; information about value added components is lacking.
  - Quarterly GDP uses a version of the SUT limited by lack of quarterly sources (Table 2); products appear only for a version of GDP consistent by production and expenditure approaches.
  - Benchmarking technique: Denton technique (P. Cholette, StatCan) used to combine annual estimates with quarterly indicators.
  - Quarterly system derives series from seasonally unadjusted source data, providing unadjusted quarterly estimates.
  - Seasonal adjustment: X-12 ARIMA is used.
- Monthly (IMACEC):
  - IMACEC is a Laspeyres volume index integrating monthly production indices of relevant industries by composition of value added for 2003 (base year).
  - Industry indices compiled monthly: agriculture; fishing; mining; manufacturing; electricity, gas and water; construction; transport and communication services; trade, restaurant and hotel services; financial and business services; social and personal services; governmental services.
  - IMACEC covers 83 percent of GDP for the base year (2003).
  - Deficiencies in coverage, especially for personal and business services.
  - Seasonal adjustment: X-12 ARIMA; moving averages are "3 x 9" for the seasonal component and a Henderson of 39 for the trend-cycle.

*Source: 3.2 Assessment of source data (National Accounts), _cr07318 - 3.2 Assessment of source data*

### 3.4 Assessment and validation of intermediate data and statistical outputs

### 3.4 Assessment and validation of intermediate data and statistical outputs

### Validation practices
- Data derived from enterprise surveys are crosschecked against:
  - output data from administrative sources,
  - merchandise trade statistics,
  - financial statements when available.
- Foreign trade information is validated with data from agriculture and mining corporations.
- Basic data cross-checked with other sources (survey and production data vs. financial statements).

### Investigation of discrepancies
- Ad hoc assessments of potential discrepancies are performed in some cases (copper, oil, and fish flour).
- Checks occur after discrepancies are identified in the supply-and-use table.
- When discrepancies in intermediate data are identified, they are evaluated and investigated with the objective of removing them.
- If business survey information is unclear or shows irregular behavior:
  - specialists call enterprises’ accountants for explanations to reconcile information,
  - respondents are sometimes visited to validate data.

### Institutional approach and limitations
- The NAD investigates any statistical discrepancies and other potential indicators of problems.
- The supply-use framework is used to make GDP estimates by activity and expenditure components consistent.
- In general, discrepancies are assumed to be part of the expenditures on GDP since production-side sources are considered more reliable.
- NAD does not evaluate data against unofficial estimates (e.g., from government units, universities, nongovernmental organizations, or international organizations) to avoid possible influence.
- NAD does not carry out studies to ensure that bias in the GDP estimate is negligible and stable over time.

---

### 3.5 Revision studies

### Routine analyses
- Direction and magnitude of revisions between preliminary data and final data are analyzed and investigated.
- When final GDP figures change with respect to preliminary estimates owing to coverage improvements, changes in trends are internally explained.
- These analyses are published in the CBCH’s Monthly Bulletin.

### Documentation
- Adequate documentation on revisions is maintained and includes:
  - description of causes of revisions,
  - methods used to incorporate new data sources,
  - how data are adjusted.

---

### 4 Serviceability

### 4.1 Periodicity and timeliness
- Periodicity:
  - Quarterly national accounts are compiled and meet the SDDS requirement on periodicity.
- Timeliness:
  - Quarterly GDP estimates are disseminated 35 days after the end of reference quarter.
  - Annual GDP estimates are disseminated within six months after the end of reference year, meeting international standards.

### 4.2 Consistency

Findings on internal consistency
- GDP estimates are produced by the production and the expenditure approaches.
- Similar and consistent growth rates are obtained from GDP by activity and GDP by expenditure category.
- GDP at current prices, volume measures, and implicit deflators are consistent within the “value = volume × price” framework.
- Discrepancies between supply and expenditures are adjusted as part of final household consumption expenditures and changes in inventories because supply sources are considered more reliable than expenditure sources.
- Quarterly estimates on construction might be overestimated due to sources (construction permits) and methods used (no adjustments for lags in project execution).
- Quarterly GDP estimates are consistent with annual estimates.

Findings on temporal consistency
- The latest major revision with base year 2003 allows coherent time series for the period 2003–2006.
- Interpolation for 1986–2002 was done by applying the splicing method to prepare comparable series with 2003 as base year.
- These series were published in staff working papers and are posted on the CBCH’s website.
- Detailed methodological notes explain breaks and discontinuities and adjustments made to maintain consistency.
- Unusual changes in economic trends are explained in analytical text included in the Monthly Bulletin of the CBCH.

Recommendation
- Reconstruct historical series as far back as reasonably possible (at least five years) when changes in source data, methodology, and statistical techniques are introduced (e.g., change of benchmark year).
- The linked series for 1986–1995 should be reviewed and the series 1986-2002 should be disseminated in a hard-copy publication.

Findings on consistency with other datasets
- NA statistics are largely consistent with balance of payments (BOP) statistics.
- Monetary statistics and government finance statistics (GFS) are reconcilable with NA estimates since the same data sources are used.
- NAD has carried out reconciliation exercises between NA and BOP and GFS; differences in coverage, sectorization, and classification are reconcilable.

### 4.3 Revision policy and practice

- Revisions follow a predetermined and reasonably stable cycle that is made known to the public in advance; reasons underlying the cycle are explained.
- Adequate documentation of revisions is included in publications and on the CBCH’s website.
- Preliminary and/or revised data are clearly identified:
  - Forecast of the percentage change of GDP at constant prices for the current year is presented in the monetary program on the CBCH’s website.
  - The first annual estimate of GDP (based on quarterly estimates) is disseminated with more detail at current and constant prices.
  - Users are informed by a footnote in each table that initially published data are preliminary and subject to revision; preliminary data are labeled.
  - Revised data are disseminated with the same level of detail as previously published for the data being revised.
- Studies and analyses of revisions are made public:
  - Analysis of differences between revised and preliminary data is published for major aggregates to allow assessment of reliability of preliminary data.
  - Analysis and causes of major revisions (e.g., change in the base year) are also made public.
  - Note: The series 2003-2006 were published in hardcopy on May 17, 2007. The revised series with 2003 as base year were disseminated on March 23, 2007.

---

### 5 Accessibility

### 5.1 Data accessibility

Presentation and detail
- NA are readily accessible on the CBCH’s website.
- Annual and quarterly NA estimates are presented clearly; charts, tables, and analysis of recent developments accompany the data.
- Household and government final consumption expenditures are disseminated at a high level of aggregation; datasets are not published with sufficient detail to facilitate meaningful comparisons.
- Recommendation: Increase the level of detail of disseminated annual data at three digits of the ISIC, or at least at two digits.

Dissemination media and formats
- IMACEC and quarterly and annual GDP are disseminated through press releases and posted on the CBCH’s website.
- More detailed data are published in hard-copy publications: the Monthly Bulletin of the CBCH and the Yearbook of National Accounts of Chile; CBCH publications are sold for a price.
- Electronic versions are available on the CBCH’s website.
- Longer time series can be accessed electronically and free of charge in the Statistical Database on the CBCH’s website.

Preannounced schedule and simultaneous release
- A twelve-month advanced calendar of economic data release announces release dates; schedule is disseminated in the CBCH’s Monthly Bulletin and on the CBCH’s website.
- Statistics are released punctually according to the preannounced schedule.
- Statistics are made available to all interested users simultaneously.
- When IMACEC and GDP estimates are provided 24 hours in advance to the MOF, embargos are imposed to prevent early public disclosure.

Access to non-routinely disseminated data
- Nonpublished, but nonconfidential subaggregates are available upon request; availability is publicized.
- A form called “Request of intermediate statistical information unofficially published by the Central Bank of Chile” must be filled out; the form is available at the CBCH library.

### 5.2 Metadata accessibility

Documentation available
- The National Accounts of Chile, 2003 Benchmark Compilation of November 2006 presents sources and methods for benchmark year 2003 and series 2003–2006.
- Updated comprehensive metadata are available in:
  - the 2003 Benchmark Compilation,
  - the Update of the Methodology of the IMACEC 2003 base,
  - Quarterly National Accounts of Chile 1996–2004: Methodology and Results.
- SDDS metadata are available on the CBCH’s website but need updating given the recent dissemination of the revised series with 2003 as base year in March 23, 2007.
- The CBCH has a project for developing a metadata system to be posted on the CBCH’s website.

Recommendation
- Update the SDDS metadata for national accounts as soon as possible.

Levels of detail
- A general brochure Síntesis Estadística de Chile 2001-2005 is available to the public.
- Working documents and background documents on the 2003 Benchmark Compilation are being prepared for public dissemination.

### 5.3 Assistance to users

Contact points and services
- Prompt and knowledgeable service and support is available at the CBCH library for publications and the electronic database.
- Two sites for queries and requests on NA statistics are available on the CBCH’s website.
- Address, phone, and fax numbers of the Publications Department are available on the CBCH’s website.
- A direct contact person for each subject field is not publicized in publications or on the website; the CBCH library receives electronic queries and forwards requests to the appropriate contact person.
- Services are publicized on the CBCH’s website and in its publications.

Catalogs and charges
- A catalog of publications, documents, and other services is available and updated each year.
- The publications catalog and presentation of CBCH products and services can be accessed on the Internet site.
- The catalog clearly indicates prices charged; assistance in placing orders is available through the CBCH’s Publication Department and an online ordering site exists on the CBCH’s website.

---

*Source: _cr07318 - 3.4 Assessment and validation of intermediate data and statistical outputs*

### 3.5              Revision              studies

### 3.5              Revision              studies

### Serviceability (Section 4)
- 4.1 Periodicity and timeliness: Practice marked with "X" in assessment matrix.
- 4.2 Consistency: Practice marked with "X" in assessment matrix.
- 4.3 Revision policy and practice: Practice marked with "X" in assessment matrix.

### Accessibility (Section 5)
- 5.1 Data accessibility: Practice marked with "X".
  - Released data are too aggregated.
  - Recommendation: Data should be disseminated at least at two digits (or preferable at three digits) of ISIC.
- 5.2 Metadata accessibility: Practice marked with "X".
- 5.3 Assistance to users: Practice marked with "X".

### Price Statistics (Consumer Price Index) — Overview (II)
- Compiling agency: National Statistics Institute (NSI).
- Legal basis: Law 17374/1970 (Article 2 establishes responsibility; Article 1 defines NSI as decentralized independent technical entity).
- Planned publication: New CPI planned to be published by December 2008.

### 0. Prerequisites of quality — Key findings and arrangements
- 0.1 Legal and institutional environment
  - 0.1.1 Responsibility for collecting, processing, and disseminating CPI clearly specified; CPI included in the National Statistics Plan.
  - 0.1.2 Data sharing and coordination: Adequate procedures; Article 20, Law 17374 requires individuals and firms to provide statistical data; National Statistics Commission (CNE) facilitates data sharing; arrangements in place to consult with Central Bank of Chile (CBCH) on new CPI.
  - 0.1.3 Confidentiality: Article 29 of Law 17374 guarantees confidentiality; breach punished by Article 247 of the Penal Code; Article 30 restricts publication referencing individuals/entities; access to individual price data restricted; unit records anonymized for analysis; only three members from upper management may enter the computer system that allows information extraction and CPI computation.
  - 0.1.4 Statistical reporting mandate: Article 20 and Article 21 make reporting compulsory; penalties for noncompliance specified in Article 22, Law 17374 (penalties referred to minimum wages and scaled regularly); penalties are sometimes low and rarely levied but act as deterrent.

- 0.2 Resources
  - 0.2.1 Staffing and facilities
    - CPI program staff total: 22.
    - Staff structure: 4 coordinators permanently verify data collected by 12 field agents; about 10,000 prices are received directly from users.
    - CPI basket: approximately 90,000 price quotes; about 80,000 are collected on a monthly basis directly at outlets by the 12 field agents.
    - Ratio of price quotes per field agent: about 6,700 quotes per month.
    - Staff mix table (Permanent / Contract / Total):
      - Department Unit Chief: 1 / 0 / 1
      - Analyst: 0 / 1 / 1
      - CPI Unit Chief: 1 / 0 / 1
      - Processors: 2 / 0 / 2
      - Coordinators: 4 / 0 / 4
      - Field Supervisor: 1 / 0 / 1
      - Field agents: 0 / 12 / 12
      - Total: 9 / 13 / 22
    - Computing resources: sufficient but outdated and not flexible enough to introduce methodological improvements.
    - Contingency/business continuity: No emergency back-up system or business continuity plan to compile CPI at alternate location.
    - Office conditions: limited space and insufficient cooling system.
    - Funding: adequate for regular needs; project funding (e.g., Household Budget Survey every five years) not secure; funding horizon restricted to a one-year period.
    - Recommendations:
      - Allocate sufficient staff and office space; form group devoted to analysis of source data, methodologies, and quality changes; increase number of field agents to expand outlet sample and avoid concentration.
      - Design a contingency plan; secure computing resources to allow CPI compilation at an alternate location in case a natural disaster or unusual event impedes access to NSI building.
  - 0.2.2 Measures to ensure efficient use of resources:
    - Management performs annual staff evaluations, periodic reviews of work processes and sample size; NSI pursues outside expert assistance.

- 0.3 Relevance
  - 0.3.1 Monitoring relevance and utility:
    - CNE conducts periodic and structured consultation processes; NSI keeps data users informed via webpage and participates in international/professional meetings.
    - Limited studies undertaken by NSI to identify new/emerging requirements.
    - Recommendation: Conduct survey among data users to identify CPI requirements focusing on data dissemination and analysis.

- 0.4 Other quality management
  - 0.4.1 Processes to focus on quality:
    - NSI strategic objectives include implementation of a quality management system; staff training emphasizes importance of quality.
    - Field agents are not continuously trained on collection techniques.
    - NSI participated in previous ROSC exercise and plans to implement ISO 9001:2000 quality management system on the CPI by December 2007.
    - Recommendation: Implement on schedule planned ISO 9001:2000 quality management system.
  - 0.4.2 Processes to monitor quality:
    - Price collection monitoring conducted, but not structured or routinely reported to management.
    - Random revisions by Price Statistics Department Chief.
    - Recommendation: Produce reports for management on price collection errors, data processing as well as product and outlet substitutions.
  - 0.4.3 Quality considerations in planning:
    - Quality improvements identified from ongoing monitoring; program lacks periodic quality reviews.
    - Feedback mainly from National Statistics Commission and large users; no surveys among general public.
    - Sample design trade-offs not adequately considered; outlet selection driven mainly by budgetary considerations without adequate statistical considerations.
    - Recommendation: Select CPI sample balancing budgetary restrictions and statistical requirements to attain an accurate representation of consumer habits.

### 1. Assurances of integrity — Key findings and recommendations
- 1.1 Professionalism
  - 1.1.1 Impartial production:
    - Law 17374 recognizes NSI independence to determine methodology, release dates, survey design, and definitions.
    - NSI Director appointed by the President from a group proposed by the Civil Service Council; deputy directors elected by the director from a proposed group.
    - Tenure: three years plus one possible reelection; short tenure and discretion for removal are not supportive of professional independence.
    - Recruitment: contest-based per Article 53 of Administrative Statute 18834 and Decree 69; legal rigidity to terminate contracts with nonperforming staff.
    - Recommendation: Strengthen independence by lengthening tenure of NSI Director and Deputy Directors from three to six years, staggering appointments, maintaining reappointment option, and limiting discretion for removal to specific causes such as breach of statistical secret or inappropriate use of confidential information.
  - 1.1.2 Choices of sources and techniques:
    - HBS source selection and statistical techniques based solely on statistical considerations.
    - Product and outlet selection influenced by budgetary restrictions.
  - 1.1.3 Right to comment on misuse:
    - NSI responds to erroneous interpretations via press releases, public appearances, media briefings; monitors media coverage regularly.

- 1.2 Transparency
  - 1.2.1 Terms and conditions public:
    - NSI website reproduces Statistics Law 17374, National Statistics Plan, calendar for CPI publication, metadata, and a contact phone number.
  - 1.2.2 Internal governmental access:
    - No government officials outside NSI have pre-release access.
    - Vulnerability: Corporate Image Department staff (outside Price Statistics Department) have access prior to release; mitigated by legal commitment signed by Corporate Image Department staff forbidding disclosure.
    - Recommendation: Reinforce CPI confidentiality by restricting pre-release access within NSI to the Price Statistics Department and upper management.
  - 1.2.3 Identification of statistical products:
    - CPI publications clearly identified with NSI name and logo; NSI requests attribution when statistics are used or reproduced.
  - 1.2.4 Advance notice of major changes:
    - Users are informed in advance of major methodological or source changes; NSI announced new HBS and new CPI to be published by end of 2008; change in public transportation computation methodology was explained via a brief document published on NSI website.

- 1.3 Ethical standards
  - 1.3.1 Staff behavior guidelines:
    - Administrative Statute Law 18834/1989 details public servants’ obligations and prohibitions (Article 84) and conflict-of-interest exclusions (Articles 85 and 86; Article 87 on teaching appointments); penalties specified in Article 121.
    - Strong culture discouraging political interference; staff sign legal pledge adhering to the “Statistical Secret.”
    - NSI publishes principles emphasizing professionalism and ethics, including that information collected is strictly confidential and used exclusively for statistical purposes.

*Source: _cr07318 - 3.5              Revision              studies (extracts from Price Statistics (Consumer Price Index) chapter of the IMF data quality assessment)*

### 2.         Methodological         soundness

### 2.         Methodological         soundness

### 2.1 Concepts and definitions
- Findings:
  - Concepts and definitions used in the CPI are generally in line with internationally accepted standards as defined by the International Labor Organization (ILO), the IMF, and others in the Consumer Price Index Manual, 2004.
  - Prices and weights broadly follow national accounts concepts based on the System of National Accounts 1993 (1993 SNA) and the European System of Accounts 1995 (1995 ESA) for all monetary household consumption expenditures.
  - Expenditure estimates are compiled at sufficient commodity and service detail for analysis at divisions, groups, classes, and items.
  - Household Budget Survey (HBS) showed expenditures on video cameras, dishwashers, and car rentals were large enough to be included in the CPI basket (above 0.025 percent) but were purposively excluded because expenditures were concentrated mainly in high-income families.
  - CPI basket includes interest rates on consumption loans and mortgage payments:
    - Nominal interest rates on consumption loans are scaled with UF investment units (double counting inappropriately inflationary effects) without disentangling capital formation components from the intermediation service component within nominal interest rates.
    - Mortgage payments include amortization charges, which is an investment rather than a consumption concept.
    - Weight of both items in the CPI basket is significant, adding to about 3.6 percent.
- Recommendations:
  - Revise definition financial expenditures and mortgage payments to represent adequately the implicit charges for the purchase of financial services and avoid capital formation components embedded within interest rates.
  - Add items with a large expenditure share from high-income families purposively excluded from the current basket.

### 2.2 Scope
- Findings:
  - The CPI measures change in prices of a fixed basket representative of final consumption expenditures of households of Greater Santiago (about 40 percent of Chile’s population), thus lacking a national perspective by excluding other cities and rural areas.
  - NSI is conducting a new national HBS and plans to update the CPI base with the expenditure results from this survey by December 2008.
  - CPI coverage includes purchases of market goods and services; excludes illegal goods and services and rental equivalence on owner-occupied housing (NSI is considering including owner-occupied housing in the new CPI).
- Recommendation:
  - Expand scope of the HBS to include expenditure patterns and price information from other regions in Chile and consider adding owner-occupied housing to the basket.

### 2.3 Classification/sectorization
- Findings:
  - CPI is based on a domestic classification closely corresponding to COICOP with departures at group levels that hinder international comparisons:
    - Food category includes Restaurants and Alcoholic Beverages (COICOP distinguishes Food and Non-Alcoholic Beverages from Alcoholic Beverages and Restaurants).
    - Tobacco is classified as Other Expenditures (COICOP suggests Tobacco be classified with Alcoholic Beverages and Tobacco).
    - Education is bundled with Recreational Services (COICOP suggests separate categories).
  - Available information allows regrouping domestic classification into COICOP.
  - Classification structure: eight groups, 41 subgroups, 156 articles, 482 products, and more than 1,300 varieties. NSI webpage publishes a list of 483 products, placing a zero weight on one of those (low octane gasoline).
  - Domestic classification splits educational services into different product payments for same service (school’s annual fee, monthly payment, parents center fee); geometric aggregation at a lower level may produce index bias that underestimates inflation in the educational sector.
- Recommendations:
  - Reclassify educational services at the product level; merge within a product category all payment types of the same service to avoid index bias.
  - Adopt COICOP classification system for CPI basket to facilitate international comparisons.
  - Exclude low octane gasoline from CPI basket (since it has a zero weight).

### 2.4 Basis for recording
- 2.4.1 Market prices are used to value flows and stocks:
  - Findings:
    - CPI weights based on expenditure data from the 1996–1997 HBS.
    - Consumption expenditure valued at purchasers’ prices including trade and transportation margins and applicable taxes (value added tax) less subsidies.
    - Product specifications include relevant price-determining characteristics.
    - Nonconditional discounts that apply on cash purchases to all consumers are considered.
- 2.4.2 Recording is done on an accrual basis:
  - Findings:
    - Prices recorded in the period they are purchased.
- 2.4.3 Grossing/netting procedures:
  - Findings:
    - Weights for used cars are not netted from disposals; HBS did not specify new vs. used car expenditures and CPI weights were obtained from alternate sources (estimating used cars weight according to stocks without appropriate netting).
    - Weight on used cars in CPI is relatively high: 1.12 percent on used cars vs. 0.67 percent on new cars.
  - Recommendation:
    - Revise methodology to derive used cars weight in CPI by adopting internationally accepted standards and considering particular conditions of the used cars market in Chile (imports are not allowed).

### 3. Accuracy and reliability

#### 3.1 Source data
- 3.1.1 Source data collection programs:
  - Findings:
    - CPI basket built with approximately 90,000 monthly price quotes (reduced from an initial 130,000, excluding items with similar dynamics).
    - Sample reduction may increase accuracy of surviving quotes but decrease variety and outlet representativeness.
    - CPI weights and basket structure based on 1996–1997 HBS. Previous surveys: 1956–1957, 1968–1969, 1977–1978, 1987-1988.
    - 1996–1997 HBS levied during a 12-month period (August-July) with seasonal variation accounted for; participating households totaled 8,445 selected with a stratified probabilistic sampling; sampling frame from 1992 Census had 1,150,947 households.
    - HBS scope restricted to Greater Santiago (Province of Santiago, Puente Alto, San Bernardo, and Padre Hurtado).
    - NSI planned updates every five years but budgetary restrictions impeded implementation; a new national HBS is currently being conducted. The 10-year cycle that prevailed is likely too infrequent given Chile’s economic growth.
    - Regular price survey collects about 80,000 specific products in the field and 10,000 provided directly by firms (product availability at retail point cannot be confirmed for firm-provided quotes).
    - Coverage excludes parallel market purchases.
    - Random sampling used for product varieties; outlets selected solely on budgetary considerations with a location criterion (six-block radius of field agents daily routes).
    - New product identification based on field agents’ day-to-day observations; insufficient staff prevents ad-hoc surveys; outlets not routinely refreshed (new outlets introduced only when previous source is out of business).
    - Periodic meetings are not held with consumer groups to identify new developments.
    - Price Statistics Department requested staff to read the Consumer Price Index Manual (2004) in Spanish to monitor developments in international standards.
  - Recommendations:
    - Plan budget in order to collect HBS every five years.
    - Analyze, refresh, and expand outlet and product sample; selection should be statistically driven (not solely based on budgetary considerations).
    - Conduct regular studies on new products and outlets to continuously refresh the sample.
    - Establish periodic meetings with consumer groups to identify new developments for CPI compilation.
- 3.1.2 Source data approximation:
  - Findings:
    - Frequency of price collection (once per week for foodstuffs and monthly for all others) is adequate for monthly CPI publication.
    - Price information consistent with definitions, scope, classifications, time of recording, and valuation required.
- 3.1.3 Timeliness:
  - Findings:
    - Periodicity and timeliness of price collection survey adequate for CPI dissemination; respondents aware of deadlines and compilers follow up for timely receipt.

#### 3.2 Assessment of source data
- Findings:
  - Various validation procedures exist for source data, but no established procedures to verify HBS response consistency with other sources.
  - International experience suggests alcoholic beverages and tobacco are usually underreported; rental equivalence on owner-occupied housing tends to be overreported.
  - Accuracy of price quotes provided directly to NSI (about 10,000 quotes per month) are routinely assessed on a random basis and typically precise.

#### 3.3 Statistical techniques
- 3.3.1 Data compilation:
  - Findings:
    - Compilation procedures minimize processing errors (coding, editing, tabulation).
    - Adjustments to individual records made only when clearly warranted; unusual values not replaced unless clearly required and identifiable.
    - Appropriate methods used to handle temporarily missing prices; product or class imputation procedures in place for nongeneralized scarcity.
    - If prices missing for less than 35 percent of varieties, a carry forward technique is applied.
    - A product with price quote missed repeatedly during three consecutive months is replaced.
    - Quality adjustments restricted to variations in weight or content capacity; other dimensions of quality change not considered except for computers (procedure needs revision as it might overestimate quality improvements leading to a downward bias).
    - New products that would modify CPI basket introduced only during a major revision (change in the weight reference period); varieties incorporated opportunistically as they gain market share.
  - Recommendations:
    - Determine procedures to adjust prices for products and services with quality changes, prioritizing procedures for large expenditure share products and items that frequently experience quality changes such as electronic devices.
    - Upgrade methodology to adjust computer prices when quality changes.
    - Discontinue use of a carry forward method in handling missing prices.
- 3.3.2 Other statistical procedures:
  - Findings:
    - Monetary household consumption expenditure, as defined in the 1993 SNA, used to establish weights; excludes illegal markets and owner-occupied housing.
    - Value recorded represents economic cost to consumer at time of purchase regardless of payment method.
    - Geometric mean formula used to calculate elementary indices:
      - Appropriate for most elements, but may produce a downward bias for educational services and other products (e.g., electricity) where varieties are not substitutable.
      - Educational services split into different products reflecting payments toward same service; geometric aggregation of non-substitutable varieties should be revised.
      - Elementary food indices aggregated with geometric mean; rest of elementary indices use Laspeyres formula.
    - Weight reference period 1996–1997 was adjusted to current December 1998 CPI base with relative price movements (i.e., the CPI is a Lowe index); this procedure is not explained in methodological documentation on NSI’s website.
    - New index was linked using internationally accepted practices when new weights were introduced (December 1998).
  - Recommendation:
    - Review products in which varieties are not substitutable and adjust elementary index aggregation formula if appropriate.

#### 3.4 Assessment and validation of intermediate data and statistical outputs
- 3.4.1 Validation against other information:
  - Findings:
    - CPI regularly compared with Producer Price Index (PPI) and the wholesale price index; plausible and widely accepted estimates are available for the covered part of the economy.
- 3.4.2 Investigation of statistical discrepancies:
  - Findings:
    - Strict validation program for price quotes in place; price quotes permanently audited by four staff members.
    - Thresholds for investigating unusual price movements are invariant across items: a uniform +/- 10 percent boundary is used to investigate price variations.
    - Mean and variance of price movements presumably differ across product categories and over time; uniform threshold may under- or over-validate results.
  - Recommendations:
    - Redefine thresholds for unusual price movements of index components; base boundaries on idiosyncratic characteristics of each item or group.
    - Review boundaries periodically to account for changes in economic trends and avoid stale thresholds.
- 3.4.3 Investigation of discrepancies and indicators:
  - Findings:
    - Price data checked for consistency within same product category and appropriate adjustments made to remove discrepancies.

#### 3.5 Revision studies
- Findings:
  - HBSs conducted on a 10-year cycle; weight revisions are not carried out periodically to determine effects of substitution bias on the CPI.
  - NSI does not study long-term trends in revision patterns to identify systematic biases within HBS revisions; misreporting on specific products (e.g., alcoholic beverages) not analyzed.
  - Errors in price data are under control but not registered (documentation inadequate), preventing analysis of error sources.
- Recommendations:
  - Analyze alternate sources of information to the HBS to track expenditure patterns intra-HBS’s.
  - Document errors and intermediate data revisions.

*Source: _cr07318 - 2.         Methodological         soundness*

### 4.         Serviceability

### 4.         Serviceability

### 4.1 Periodicity and timeliness
- Findings:
  - The CPI is published on a monthly basis.
  - The periodicity of the CPI conforms to the IMF’s Special Data Dissemination Standard (SDDS).
  - Publication of the index is within the first five working days after the end of the reference month (while the requirement is to publish within one month).
  - The CPI timeliness exceeds the IMF’s SDDS.
- Recommendations:
  - (No additional specific recommendation in this subsection beyond conformity and timeliness observations.)

### 4.2 Consistency
- 4.2.1 Statistics are consistent within the dataset
  - Findings:
    - The statistical series are internally consistent.
    - The Chilean CPI has a unique classification typology item/expenditure class since the scope is restricted to Greater Santiago.
- 4.2.2 Statistics are consistent or reconcilable over a reasonable period of time
  - Findings:
    - Consistent time series for the CPI’s general index is available for the period starting on January 1960 on the NSI’s webpage.
    - Historical series are reconstructed, as far as possible, when changes in source data, methodology, and statistical techniques are introduced.
    - Price indices at the product level are available only from January 1999 on the NSI’s webpage (data from previous periods are available upon request).
    - Methodological notes identify and explain (scantly) the main breaks and discontinuities in time series, their causes, as well as adjustments made to maintain consistency over time.
    - The press release (or any other NSI publication) does not provide an explanation of unusual changes in economic trends.
  - Recommendations:
    - Improve analysis on press release to facilitate proper interpretation. In particular, explain causes of contemporaneous CPI movements.
    - Provide more detail in methodological notes that explain main breaks in CPI time series.
- 4.2.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks
  - Findings:
    - Consumer price statistics are largely consistent with producer price index (excluding copper) as well as with the wholesale price index.

### 4.3 Revision policy and practice
- 4.3.1 Revisions follow a regular and transparent schedule
  - Findings:
    - The CPI is final upon its first release, and this is common knowledge among users.
    - Use of the CPI in legal contracts and in determining investment units (UF) impedes revisions.
- 4.3.2 Preliminary and/or revised data are clearly identified
  - Findings:
    - The CPI is final upon its first release; there are no preliminary CPI data.
- 4.3.3 Studies and analyses of revisions are made public
  - Findings:
    - The CPI is final upon its first release, hence no revision studies are conducted.

### 5. Accessibility

#### 5.1 Data accessibility
- 5.1.1 Presentation, layout, and clarity
  - Findings:
    - The CPI presentation is not commensurate with users’ needs.
    - A monthly press release is published by the NSI, but the charts and tables included do not facilitate a good understanding of inflationary trends.
    - Relevant graphs and tables are not always considered (e.g., an incidence CPI breakdown table).
    - Graphs have a short span (12 months) that does not portray medium-term CPI trends.
    - The introductory summary in the release does not contain pertinent comparisons (annual variations are not reported in some cases; otherwise comparisons are not made with the results from the previous month).
    - The press release does not offer meaningful analysis on contemporaneous price developments.
    - The press release does not report on its first page core inflation results either within the text, graphs, or tables (it is usually reported on the fourth page).
    - The NSI does not report the CBCH core inflation measure X1; the CPI X1 indicator is produced by the CBCH instead of the NSI, delaying public knowledge and hindering understanding of medium-term inflationary trends.
    - CPI statistics are not disseminated in a seasonally adjusted form.
  - Recommendations:
    - Improve presentation and content of press release to facilitate proper interpretation. In particular, the press release will benefit from further analysis, systematic inclusion of relevant charts and graphs, use of pertinent reference points, and highlighting core inflation results (considering both the NSI and CBCH X1 definitions).
    - Develop a more friendly webpage. In particular, publish X1 price index (CBCH core definition).
- 5.1.2 Dissemination media and format
  - Findings:
    - The CPI dissemination format could suit users’ needs better.
    - Although CPI data are split between tradables and nontradables, the NSI does not publish core measures of tradables and nontradables.
    - CPI statistics are published on the NSI’s webpage with detail starting in January 1999, while longer series are available upon request.
    - Detailed time series are separated by calendar year; both procedures do not facilitate economic analysis.
  - Recommendations:
    - Publish core tradables and nontradables measures (with NSI and X1 CBCH definitions).
    - Enhance access to historical time series at the product level.
- 5.1.3 Release calendar
  - Findings:
    - The NSI publishes a calendar which announces twelve months in advance the CPI release dates on its webpage.
    - On the preannounced date at 8:00 a.m., statistics are punctually released.
- 5.1.4 Simultaneous availability
  - Findings:
    - The CPI is available to all users at the same time.
    - The public is aware of access procedures (NSI webpage) which offer simultaneous availability of the statistics.
    - The press is not briefed in advance.
- 5.1.5 Non-routinely disseminated statistics
  - Findings:
    - Statistics not routinely disseminated are available upon request.
    - Customized tabulations can be provided with a fee (determined by an expert who estimates the amount of time required to process the request).
    - Procedures to obtain additional information are made known to the public through the NSI’s consultation desk. The desk may be contacted by phone. A list of several phone numbers is reported on the NSI’s webpage with a specific contact person for each statistic.
  - Recommendations:
    - Add an e-mail address for the consultation desk, in order to better track response time and quality content.
    - Establish a unique phone number for consultations and route to pertinent specialist, in order to avoid nonresponse during absence of contact referred in the list.

#### 5.2 Metadata accessibility
- 5.2.1 Documentation on concepts, scope, classifications, sources, techniques, and departures from standards
  - Findings:
    - The NSI publishes on its website methodological documents for the HBS and the CPI compilation.
    - In the case of the HBS, it would be useful to include the questionnaire that was used.
    - The methodological document includes most relevant methodological aspects of the CPI compilation (concepts, definitions, classifications, data sources, compilation methods, and statistical techniques).
    - The document has not been updated with changes the CPI has experienced during the last decade (e.g., the number of price quotes and outlets in the sample has decreased since its publication).
    - The methodological document does not explain with enough detail issues on the construction of the index weight structure (such as relative price updating of HBS results or the split between used and new cars from the generic “cars” expenditure item).
    - Departures from internationally accepted standards in the CPI are not explained in the methodological document (e.g., differences with COICOP).
    - SDDS metadata and summary are reviewed regularly.
  - Recommendations:
    - Upgrade CPI methodological document and update it periodically.
    - Publish the HBS questionnaire on NSI’s website.
- 5.2.2 Levels of detail adapted to audience
  - Findings:
    - General use information about the CPI is scant on the NSI’s webpage.
    - The methodological document is geared mainly toward a specialist audience.
    - Documents that explain methodological changes are specialist oriented (example: the change on the compilation methodology on public transportation was difficult to understand by the public).
  - Recommendation:
    - Publish general use information about the CPI on the NSI’s webpage (explain what is the CPI, what it is used for, and how it is built).

#### 5.3 Assistance to users
- 5.3.1 Contact points
  - Findings:
    - Prompt and knowledgeable service and support is available for statistics users.
    - Press releases include the NSI address, telephone number, and the e-mail from the press division.
    - The contact point for the consultation desk is not provided.
    - Assistance to users is monitored by the consultation department mainly on the timeliness of responses to requests, but the content of these are not routinely assessed by the Price Statistics Department.
  - Recommendation:
    - Assess periodically quality content of responses offered by the consultation desk by the Price Statistics Department.
- 5.3.2 Catalogs and availability of publications
  - Findings:
    - A catalog of NSI publications is available to users and regularly updated.
    - The catalog discloses clearly the prices of statistical products.
    - Products may be purchased electronically or on the NSI’s premises.

### Key statistics and quality observations (from Data Quality Assessment Framework)
- The CPI program lacks adequate staff, facilities, and computing resources; computers and compilation software are outdated; business continuity is not guaranteed.
- The CPI basket includes interest rates on consumption loans and mortgage payments, with a treatment inconsistent with best practices (nominal interest rates scaled with inflation not distinguishing capital accumulation components from consumption components).
- The CPI scope is limited to Greater Santiago (about 40 percent of Chile’s population); a new national HBS is being conducted and there are plans to update the CPI base with the expenditure results from this survey by December 2008.
- Domestic classification departs from COICOP in some respects (e.g., Food includes Restaurants and Alcoholic Beverages; Education bundled with Recreational Services), hindering international comparisons.
- Erroneous netting procedures bias the CPI weight structure: weight on used cars is 1.12 percent vs. 0.67 percent on new cars.
- Source data and sampling:
  - The CPI basket is currently built with approximately 90,000 monthly price quotes, which are obtained from about 2,200 outlets.
  - Price quotes have been reduced from an initial number of 130,000.
  - HBSs are conducted on a 10-year cycle; weight revisions are not carried out periodically to determine effects of substitution bias on the CPI.
- Statistical techniques and validation:
  - Data compilation procedures are sound and appropriate statistical methods are used to handle missing prices.
  - New products are introduced in the sample as they gain market share, but a similar treatment for outlets is lacking.
  - Adjustments for quality differences are not made, except for computers (procedure requires revision), potentially leading to inappropriate introduction of new products and services.
  - There are validation procedures for source data, but no established procedures to verify HBS response consistency with other sources.
- Revisions and studies:
  - HBSs on a 10-year cycle lead to no periodic weight revisions and no long-term trends studies for revision patterns.
  - Errors in the price data are under control but these are not registered.

*Source: _cr07318 - 4.         Serviceability*

### 4.     Serviceability     4.1   Periodicity and ti

### _cr07318 - 4.     Serviceability     4.1   Periodicity and ti

### 4. Serviceability — periodicity, timeliness, consistency, revision
- 4.1 Periodicity and timeliness: X
- 4.2 Consistency: X
- 4.3 Revision policy and practice: X

### 5. Accessibility
- 5.1 Data accessibility: X
  - CPI statistics are punctually published on a preannounced schedule.
  - The statistics are not presented in a way that facilitates proper interpretation and meaningful comparisons.
  - Appropriate analysis of current-period developments is not included in the press release.
  - Data dissemination formats hinder redissemination and analysis of inflationary trends.
- 5.2 Metadata accessibility: X
- 5.3 Assistance to users: X

### III. PRICE STATISTICS (PRODUCER PRICE INDEX) — 0. Prerequisites of quality

0.1 Legal and institutional environment
- 0.1.1 Responsibility for collecting, processing, and disseminating the statistics
  - The producer price index (PPI) is compiled by the National Statistics Institute (NSI).
  - Responsibility is clearly established in Law 17374/1970 (Article 2). The Law states that the NSI is a decentralized independent technical entity responsible for official statistics and censuses within Chile (Article 1).
  - The National Statistics Plan includes the PPI as an official statistic.
- 0.1.2 Data sharing and coordination among data-producing agencies
  - Adequate procedures in place for timely flow of producer prices collected directly from firms.
  - Article 20, Law 17374, makes reporting compulsory for individuals and firms to provide statistical data.
  - National Statistics Commission (CNE) reviews statistics periodically; CNE is composed of NSI Director, representatives from public entities including Central Bank of Chile (CBCH), Ministry of Finance (MOF), Development Corporation (CORFO), universities, private sector, and labor unions.
  - No arrangement exists to facilitate consultation with CBCH to study emerging requirements on the PPI.
  - Recommendation: Establish periodic meetings with CBCH to facilitate consultations.
- 0.1.3 Confidentiality of individual reporters’ data
  - Article 29 of Law 17374: NSI, fiscal and semi-fiscal institutions, state firms, and staff cannot disseminate information from individuals or firms learned in professional activities; “Statistical Secret” and dissemination reserves; breach is a crime punished by Article 247 of the Penal Code.
  - Article 30: statistical data cannot be published with express reference to individuals or entities if prohibited by affected parties.
  - Access to individual price data restricted to staff who require it; unit records anonymized for further analysis; only three members from upper management may access the computer system that allows information extraction and PPI computation.
  - NSI posts principle: “The information that is collected by statistical entities, from individuals and firms, is strictly confidential and should be used exclusively for statistical purposes.”
- 0.1.4 Statistical reporting ensured through legal mandate
  - Law 17374 confers legal authority to NSI to collect PPI data.
  - Article 20: reporting compulsory to private sector; Article 21 extends to public sector; Article 22 specifies penalties referenced to minimum wages.
  - Penalties sometimes low and rarely levied but act as deterrent.
  - Data collection modes: about 60 percent of the prices are collected via e-mail, 20 percent by fax, and the remaining 20 percent by direct visits to firms.

0.2 Resources
- 0.2.1 Staff, facilities, computing resources, financing
  - PPI program lacks adequate staff to perform all required tasks.
  - Staff assigned to PPI program totals five (including the Statistics Department chief).
  - PPI basket built with approximately 4,000 specific price quotes levied on 738 firms.
  - Field agents must collect, capture, review, and analyze price quotes; chief of PPI unit also responsible for wholesale price index.
  - Staff mix table (Price Statistics Department (PPI)):
    - Department Unit
    - Chief: Permanent 1 0 1
    - PPI Unit Chief: 1 0 1
    - Field agents: Permanent 0 Contract 3 Total 3
    - Total: Permanent 2 Contract 3 Total 5
  - Computing resources sufficient though computers are outdated; compilation software up to date and flexible.
  - Protection for computer resources inadequate; no business continuity plan; no emergency back-up system to compile PPI at alternate location.
  - Office building inadequate: limited space, no cooling system, only one telephone line for three field agents.
  - Funding for regular needs adequate; no funding for multi-year improvement projects; funding horizon restricted to a one-year period.
  - Recommendations:
    - Allocate sufficient staff, office space and adequate furniture. Specifically, form a group of professionals devoted to the analysis of source data, methodologies, and quality changes in products and services.
    - Increase the number of field agents to expand the sample and increase control over nonsampling errors.
    - Provide a cooling system and two additional telephone lines.
    - Design a contingency plan. Secure computing resources in order to have the ability to compile PPI at an alternate location in case a natural disaster or unusual event impedes access to the NSI building.
- 0.2.2 Measures to ensure efficient use of resources
  - Management ensures efficient use of resources; annual staff evaluations; periodic reviews of work processes; NSI pursues outside expert assistance.

0.3 Relevance
- 0.3.1 Monitoring relevance and practical utility
  - CNE conducts periodic structured consultations to review emerging data requirements.
  - NSI informs users via webpage; participates in international/professional statistical meetings.
  - Scant studies to identify new requirements; no data user surveys conducted.
  - Public knowledge on PPI limited; index credibility not well established; low usage as deflator for national accounts.
  - Possible causes: users’ inertia using wholesale price index and outdated laws referencing wholesale index instead of PPI.
  - Recommendation: Conduct survey among data users to identify reasons for the PPI’s low credibility, and implement measures to address this issue.

0.4 Other quality management
- 0.4.1 Processes to focus on quality
  - NSI strategic objectives include implementing a quality management system focused on statistical processes.
  - Staff training emphasizes quality; permanent staff participates in forums on price statistics; field agents not continuously trained on collection techniques.
  - NSI participated in a previous ROSC exercise and plans to implement ISO 9001:2000 quality management system on the PPI by December 2008.
  - Recommendation: Implement on schedule planned ISO 9001:2000 quality management system.
- 0.4.2 Processes to monitor quality of statistical program
  - Quality monitoring on price collection is conducted but not reported to management in structured routine reports.
  - Permanent staff trained on quality management systems.
  - NSI Director delegated to Technical Subdirection periodic reviews to identify steps to maintain and improve PPI quality.
  - Recommendation: Produce reports for management on price collection errors, data processing, and product and outlet substitutions.
- 0.4.3 Processes to deal with quality in planning
  - Quality improvements reported by participating firms.
  - Source data selected based on measurement objectives, data requirements, and budget restrictions.
  - Feedback mainly from CNE and large users; no surveys among the general public.

### 1. Assurances of integrity

1.1 Professionalism
- 1.1.1 Statistics produced on an impartial basis
  - Law 17374 recognizes NSI independence to determine methodology, release dates, survey design, and definitions.
  - Sample selection and data gathering are objective and impartial.
  - NSI Director appointed by the President from a group proposed by the Civil Service Council; Deputy Directors elected by the Director from a proposed group.
  - Tenure for both is three years plus one possible reelection; short tenure and discretion for removal not supportive of professional independence.
  - Recruitment based on aptitude; Article 53 of Administrative Statute 18834 and Decree 69 mandate contest-based promotions; legal rigidity to terminate contracts with nonperforming staff.
  - Recommendation: Strengthen independence of the NSI by lengthening tenure of the NSI Director and Deputy Directors from three to six years, staggering appointments, maintaining the reappointment option, and limiting discretion for removal to specific causes such as breach of statistical secret or inappropriate use of confidential information.
- 1.1.2 Choices of sources and techniques informed solely by statistical considerations
  - Source data and statistical methods based on data requirements, measurement objectives, and budgetary considerations.
  - Methodology dissemination decisions, timing, and media are based solely on statistical grounds.
- 1.1.3 Right to comment on erroneous interpretation and misuse
  - NSI would respond to erroneous interpretations through press releases, public appearances, and briefings.
  - PPI not frequently scrutinized by public; NSI monitors media coverage regularly.

1.2 Transparency
- 1.2.1 Terms and conditions publicly available
  - NSI website reproduces Law 17374, National Statistics Plan, metadata, contact phone number, and a calendar for PPI publication.
- 1.2.2 Internal governmental access prior to release
  - No government officials outside NSI have access before public release.
  - Staff from the Corporate Image Department (outside Price Statistics Department) has access prior to release; mitigated by legal commitment signed by Corporate Image staff forbidding release of confidential information.
  - Recommendation: Reinforce PPI confidentiality. Prior to data release restrict access from units within the NSI outside the Price Statistics Department and upper management.
- 1.2.3 Products clearly identified
  - PPI publications clearly identified with NSI name and logo; NSI requests attribution when statistics are used or reproduced.
- 1.2.4 Advance notice of major changes
  - Users are made aware in advance of major changes in methodology, source data, and statistical techniques.

1.3 Ethical standards
- 1.3.1 Guidelines for staff behavior
  - Administrative Statute Law 18834/1989 details public servants’ obligations and prohibitions (Article 84) and forbids specific activities to avoid conflict of interest (Article 85 and 86) with exclusions and limitations (Article 87); penalties specified (Article 121).
  - Strong culture to maintain ethical standards discourages political interference.
  - Staff mentors aware of ethical standards; old and new NSI staff sign legal pledge adhering to the “Statistical Secret.”
  - NSI publishes principles emphasizing professionalism and ethics: decisions strictly on professional grounds embedded in scientific and ethical principles for data collection, compilation, storage, and dissemination.

### 2. Methodological soundness

2.1 Concepts and definitions
- 2.1.1 Overall structure follows international standards
  - Concepts and definitions broadly conform to System of National Accounts 1993 (1993 SNA) and Producer Price Index Manual (2004).
  - Output estimates compiled at sufficient industrial and commodity detail.
  - PPI classifies industry with ISIC rev. 3 at four-digit level.
  - Commodity detail classified with CPC ver. 1.0 at five-digit level.

2.2 Scope
- 2.2.1 Scope broadly consistent with international standards
  - PPI includes six economic categories: agriculture, livestock and forestry; mining; fishing; manufacturing; electricity, gas, and water; and construction.
  - Index does not cover services such as telecommunication, transportation, financial services, tourism, and retail activity.
  - Monthly surveys collect prices for approximately 4,000 specific goods in 738 establishments.
  - Information from illegal markets and the army is not considered within the scope.
  - Recommendation: Consider covering services in a future PPI expansion.

*Source: _cr07318 - 4.     Serviceability     4.1   Periodicity and ti*

### 2.3       Classification/sectorization

### 2.3       Classification/sectorization

### Classification and sectorization findings
- Institutional units and transactions are classified according to the 1993 SNA.
- ISIC rev. 3 is used to classify economic activities of establishments and enterprises.
- Products are classified with the CPC, ver. 1.0.
- Exception: classification of copper extraction and some copper manufacturing products (such as electrolytic and cathode copper rolls) departs from international classifications but conforms with CBCH national accounts.

### Basis for recording — valuation and timing
- Market producer prices are used to value PPI weights, consistent with best practice and the 1993 SNA.
- Exception: some construction items are estimated using indices from mining and forestry PPI, remuneration indicators, and the CPI; these are introduced into cost functions consistent with CBCH practice.
- Recommendation: Consider obtaining price quotes for different construction structures directly from contractors; alternatively, review periodically construction cost functions.

- Recording is done on an accrual basis in general, with output recorded in the period produced per the 1993 SNA.
- Exception: copper prices are collected from specific transaction contracts referenced to an international price and adjusted after product delivery according to metal quality, producing a recording lag.
  - Copper PPI weight is 13.3 percent.
- Recommendation: Review methodology for copper price registries; consider registering international reference prices to achieve timelier accrual of price variations.

- Grossing/netting: Intermediate goods PPI net weights follow 1993 SNA procedures. Transactions between establishments are not recorded.

### Recommendations (recording)
- Obtain construction price quotes directly from contractors or review construction cost functions periodically.
- Review methodology for copper price registries and consider registering international reference prices.

---

### 3. Accuracy and reliability

### 3.1 Source data — coverage and timeliness
- Source data are obtained from comprehensive data collection programs approximating required definitions, scope, classifications, valuation, and time of recording.
- Sample excludes firms with less than 10 employees (not covered in the business registry), enterprises within free zones, and household unincorporated enterprises.
- Export and domestic prices are considered but are not weighted according to sales.
- The IPP base is December 2003.
- The weight structure is based on production for the years 2000–2002.
- Classification between intermediate and final goods is based on the CBCH 1996 input-output matrix.
- Adequate procedures exist to delete dead business units or disappearing products; substitutions use industrial census (two-year lag), yellow pages, or web information.
- No established procedure to continuously refresh the sample of surviving firms and products; renewal guided mainly by firm-supplied information.
- Questionnaires are pilot tested; periodic reviews and observation studies are conducted.
- Output coverage: products in each economic sector selected to cover at least 80 percent of production value; some low-production goods included for input/consumption relevance.
- Sample design for firms: probabilistic proportional to size; product selections within firms focus on most sold items.
- Data detail sufficient to derive PPIs at a four-digit industry level and five-digit product level for sectors in scope.

Recommendations:
- Implement procedure to continuously refresh surviving firms and products within the sample.
- Consider future expansion of PPI to include firms of less than 10 employees, enterprises within free zones, and household unincorporated enterprises.
- Weight export and domestic prices according to sales.

### 3.1.2 Source data approximation
- Output data from the 2000–2002 framework is sufficiently detailed for PPI program requirements, except for copper registries (see 2.4.2).

### 3.1.3 Timeliness of source data
- Periodicity and timeliness of price collection generally adequate for PPI dissemination.
- Copper price registries may experience significant within-month variation; prices are recorded once a month among producing firms, but the survey levying period omits at least the first and last week of the month.
- Construction component data from other sources are received timely.
- Respondents are aware of reporting deadlines; about 80 percent of price quotes are received by e-mail or fax while 20 percent are collected in situ. Lack of telephone lines hinders reminders.
- Recommendation: Increase recollection frequency of copper price quotes. Daily recording is attainable if international market prices are used (copper contract sales are stipulated in relation to an international reference price).

### 3.2 Assessment of source data
- Nonsampling information is well tracked. Starting on the last third of each month, nonresponse, misclassification, and survey operations are verified daily by the PPI Unit Chief and reported to the Price Statistics Department Chief.
- Sampling errors are monitored monthly by establishing confidence intervals at the specific product level.
- Extreme price variations are identified, confirmed with respondents, and records maintained. Outliers are defined as price increments or decrements which exceed 15 percent.
  - Note: thresholds may under- or overvalidate changes since price variance differs across sectors and products.
- PPI price surveys and administrative data are randomly audited by the Price Statistics Department Chief.
- Validation procedures check temporal consistency and cross-check results with other datasets.
- Excluded economic activities within the PPI (services) exceed 35 percent of Chile’s total output; implicit deflators in national accounts are available for sectors not covered.
- Recommendation: Redefine thresholds for unusual price movements; base investigation boundaries on characteristics of each product or sector and review periodically.

### 3.3 Statistical techniques
- Compilation procedures minimize processing errors; unusual values are replaced only when clearly warranted and identifiable in datasets.
- Imputation and adjustment for nonresponse are soundly based. A product imputation technique is used when nonresponse ratio for a variety is higher than 35 percent.
- Missing prices in agriculture, livestock, and forestry are handled with product imputation; missing prices elsewhere are carried forward from last observed price.
- Products permanently unavailable are replaced with a twin product.
- Adjustments for quality differences are not made; a product experiencing a quality change is replaced with the new product. This may lead to downward index bias by implicitly assuming price changes on new model introduction are due exclusively to characteristics change.
- Copper price recollection procedure increases quality-induced volatility because contracts adjust international price with metal quality variations.

Recommendations:
- Determine procedures to adjust prices for products with quality changes.
- Revise existing procedure to collect copper prices.
- Discontinue the use of the carry forward imputation technique.

### 3.3.2 Aggregation and index methodology
- PPI weights established with gross output measures as defined in the 1993 SNA.
- Owner-occupied dwellings and army output are excluded from PPI scope.
- Lower level indices aggregated with a geometric mean formula; upper level aggregation uses an arithmetic mean formula.
- Current weight reference period (2000/2002) aligned to the price reference period (April 2003); PPI may be described as a Lowe index.
- Although base is relatively up to date (April 2003), there are no plans to refresh basket or weight structure in the near future.

### 3.4 Validation of intermediate data and outputs
- Intermediate PPI results validated against wholesale price index and the CPI.
- Unusual price movements are investigated but not explained in the press release, hindering public understanding.
  - Example: annual PPI variation fluctuated between 24.2 and 6.5 percent during the past 12 months, while CPI inflation moved within the range of 2.1 and 3.9 percent.
- Recommendation: Explain unusual PPI price variations in the press release; publish a PPI measure excluding copper to improve public understanding.

- The PPI unit assesses statistical discrepancies on missing data imputation and makes appropriate adjustments.

### 3.5 Revision studies
- No periodic weight revisions are analyzed to determine effects of substitution bias on the PPI.
- Recent copper price increases suggest possible shifts in production inducing substitution bias.
- Sources of errors are investigated but not studied to improve revision methods.
- The Price Statistics Department does not conduct studies on optimal revision cycle for the PPI; updating tradition less established than for CPI (the wholesale price index predecessor has base period dating from 1992).

Recommendations:
- Study PPI weight structure to estimate effects of recent copper price increments on productive structure.
- Study revision cycle and commit to a date for PPI update.

---

### 4. Serviceability

### Periodicity and timeliness
- Periodicity: PPI published monthly; conforms to IMF SDDS.
- Timeliness: Publication within the first five working days after the end of the reference month (requirement is to publish within one month). PPI timeliness exceeds IMF SDDS.

### Consistency
- PPI is internally consistent by industry and product.
- Consistent monthly time series available since start of PPI publication (April 2003). Detailed methodological notes identify main breaks and discontinuities.
- Unusual changes in economic trends are not explained in the press release.
- PPI is largely consistent with other price statistics and national accounts.

### Revision policy and practice
- The PPI is final upon its first release.
- No predetermined revision cycle for weight structure and PPI basket; no explanations provided for indeterminacy.
- Preliminary and revised data: PPI is final upon first release.
- Studies and analyses of revisions are not made public; PPI is final upon first release.

---

### 5. Accessibility

### Presentation and dissemination
- Monthly press release is published but charts and tables do not facilitate good understanding of inflationary trends:
  - Index level graphs presented instead of annual variations.
  - Graphs have a short span (12 months), not portraying medium-term trends.
  - Introductory summary in release does not contain PPI results.
  - Press release lacks meaningful analysis on contemporaneous price developments.
  - Press release does not report a PPI measure excluding copper, which explains most PPI volatility.

Recommendation: Improve presentation and content of press release — include further analysis, systematic relevant charts and graphs, pertinent reference points, and highlight a PPI measure that excludes copper.

- Dissemination format: detailed PPI time series are not published on NSI’s website (available upon request).
  - Recommendation: Publish detailed PPI historical time series on the NSI’s website and develop a more user-friendly website.

- Release schedule: NSI publishes a calendar announcing twelve months in advance the PPI release dates on its webpage; statistics are released punctually on the preannounced date at 8:00 a.m.
- Simultaneous availability: PPI is available to all users at the same time; press is not briefed in advance.
- Non-routinely disseminated statistics: available upon request; customized tabulations can be provided with a fee assessed by an expert estimating processing time. Procedures publicized via NSI consultation desk with phone contacts listed on NSI’s webpage.

Recommendations:
- Add an e-mail address for the consultation desk to better track response time and quality.
- Establish a unique phone number for consultations and route to pertinent specialist to avoid nonresponse during absence of listed contacts.

*Source: _cr07318 - 2.3       Classification/sectorization*

### 5.2       Metadata       accessibility

### 5.2       Metadata       accessibility

### 5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques
- The NSI publishes on its website a summary of its methodological document for PPI compilation.  
- A more comprehensive document is available upon request (although there is no mention of its existence).  
- The methodological summary document does not explain with enough detail relevant information on statistical techniques, definitions, and data sources.  
- Departures from international accepted standards are not covered in the aforementioned document.  
- The document lacks survey information such as response rates, survey monitoring, and studies on nonsampling errors.  
- SDDS metadata and summary are reviewed and updated regularly.  
- Recommendation: Publish full-fledged upgraded version of PPI methodological document. Update it periodically.

### 5.2.2 Levels of detail are adapted to the needs of the intended audience
- General use information about the PPI is not available on the NSI’s webpage.  
- There is a need for a document that explains the differences between the PPI and its predecessor, the wholesale price index (which although outdated is more widely known and used).  
- The methodological document, which is available, is oriented mainly toward a specialist audience.  
- Recommendations:
  - Publish general use information about the PPI on the NSI’s webpage (explain what the PPI is, what it is used for, and how it is built).
  - Publish a document explaining the methodological differences between the PPI and wholesale price index.

### 5.3 Assistance to users (related accessibility points)
- 5.3.1 Contact points for each subject field are publicized:
  - Press releases include the NSI address, telephone number, and the e-mail from the Press Division.
  - The contact point for the consultation desk is not provided.
  - Assistance for users is monitored by the Consultation Department, mainly on the timeliness of responses to requests, but the content of these are not routinely assessed by the Price Statistics Department.
- 5.3.2 Catalogs of publications, documents, and other services:
  - A catalog of NSI publications is available to users and is regularly updated.
  - The catalog discloses clearly the prices of statistical products.
  - Products may be purchased electronically or on the NSI’s premises.

### Data Quality Assessment Framework — Producer Price Index (summary observations)
- Prerequisites of quality:
  - 0.1 Legal and institutional environment:
    - The NSI operates under a legal framework that is adequate. Legislation supporting PPI compilation is established in Law 17374 (1970).
    - Law 17374 guarantees the confidentiality of individual reporters’ data.
    - Statistical reporting is mandatory according to the law (in the case of noncompliance penalties apply).
    - There are no consultation arrangements with the CBCH.
  - 0.2 Resources:
    - The PPI program lacks adequate staff, facilities, and computing resources.
    - Insufficient staff impedes continuous analysis and implementation of methodological improvements and appropriate quality adjustments to price statistics.
    - Computers and compilation software are outdated.
    - Business continuity is not guaranteed in the contingency of no access to NSI facilities.
    - The office building does not provide adequate working facilities: space is limited, there is no cooling system, and there is only one telephone line for the three field agents.
  - 0.3 Relevance:
    - The CNE conducts periodic and structured consultation processes to review emerging data requirements.
    - The NSI keeps data users informed of specific aspects of the PPI through its webpage and participates in international and professional statistical meetings.
    - Public knowledge on the PPI is scant and index credibility has not been well established.
  - 0.4 Other quality management:
    - Assessed as Largely Not Observed (comments indicate weaknesses).

- Assurances of integrity:
  - 1.1 Professionalism:
    - Law 17374 recognizes the independence of the NSI to determine methodology, release dates, design of surveys, and definitions used.
    - The NSI Director is appointed by the President from a group proposed by the Civil Service Council. Deputy Directors are elected by the Director from a group proposed by the aforementioned Council.
    - Short tenure (three years plus one possible reelection) and discretion for removal are not supportive of professional independence.
    - Source data and statistical methods are based on data requirements, measurement objectives, and budgetary considerations.
  - 1.2 Transparency: assessed as observed.
  - 1.3 Ethical standards: assessed as observed.

- Methodological soundness:
  - 2.1 Concepts and definitions: assessed as observed.
  - 2.2 Scope: assessed as observed.
  - 2.3 Classification/sectorization:
    - Classification and sectorization used in the PPI are in broad conformity with international recommended systems.
    - Institutional units and transactions are classified according to the 1993 SNA.
    - ISIC rev. 3 is used to classify economic activities of establishments and enterprises.
    - Products are classified with the CPC ver. 1.0.
    - Exception: classification of copper extraction and some copper manufacturing products.
  - 2.4 Basis for recording:
    - PPI weights are valued at market producer prices; intermediate goods PPI weights are in accordance with the 1993 SNA.
    - Exception: some construction items are estimated using items from mining and forestry PPI indices, remuneration indices, and the CPI; these are introduced into cost functions consistent with those used by the CBCH.
    - Timing rules generally follow the 1993 SNA; output is recorded in the period produced.
    - Survey design does not allow for timely registry of copper prices because prices are collected from specific transaction contracts referenced to an international price adjusted after product delivery according to metal quality.

- Accuracy and reliability:
  - 3.1 Source data: assessed as observed.
  - 3.2 Assessment of source data: assessed as observed.
  - 3.3 Statistical techniques:
    - Sound imputation techniques are used to handle missing prices in agriculture, livestock, and forestry.
    - In other PPI economic categories, missing prices are carried forward, a practice that should be used less frequently.
    - Adjustments for quality differences are not made, potentially allowing inappropriate introduction of new products and services.
    - Copper price recollection procedure increases quality-induced volatility since price quotes are derived directly from contracts that adjust the international price with quality variations of the metal.
  - 3.4 Assessment and validation of intermediate data and statistical outputs: assessed as observed.
  - 3.5 Revision studies:
    - Sources of errors are not routinely investigated.

- Serviceability:
  - 4.1 Periodicity and timeliness: assessed as observed.
  - 4.2 Consistency: assessed as observed.
  - 4.3 Revision policy and practice: assessed as observed.

- Accessibility:
  - 5.1 Data accessibility:
    - PPI statistics are punctually published on a preannounced schedule.
    - The statistics are not presented in a way that facilitates proper interpretation and meaningful comparisons.
    - Appropriate analysis of current-period developments is not included in the press release.
    - Data dissemination formats hinder redissemination and analysis of inflationary trends.
    - A PPI indicator that excludes copper effects on its volatility is not published.
    - Detailed product price series are not available on the NSI’s website.
  - 5.2 Metadata accessibility: assessed as Largely Not Observed in parts (see 5.2.1 and 5.2.2 findings).
  - 5.3 Assistance to users: assessed as observed, but with gaps in consultation desk contact details and content assessment.

*Source: _cr07318 - 5.2       Metadata       accessibility*

### 0.2       Resources

### _cr07318 - 0.2       Resources

### 0.2 Resources — Staffing, facilities, computing resources, and financing
- The number of staff is "barely adequate" to perform tasks associated with current commitments.
- In the medium term, additional human resources may be necessary to complete the migration to the Government Finance Statistics Manual (GFSM 2001) methodology.
- Staff is partially devoted to the compilation and dissemination of GFS.
- GFS are disseminated by two different departments in the MOF: the DIPRES, and the Sub-Secretariat of Finance.
- Within the DIPRES three sectors are responsible for compiling GFS:
  - (1) the statistics sector compiles annual GFS, and is comprised of one deputy chief and two analysts,51
  - (2) the coordination sector compiles monthly and quarterly GFS, and is comprised of one deputy chief and one analyst,52
  - (3) the financial programming sector compiles GFS on central government gross debt, and is comprised of one deputy chief and one analyst partially devoted to this task.
- The Sub-Secretariat of Finance disseminates GFS on public sector gross and net debt.
- Only one analyst is responsible for consolidating GFS compiled by the DIPRES, and other data-producing agencies, to report directly to the minister.
- Staff qualifications are adequate; senior staff in charge of compilation have participated in several courses based on GFSM 2001 methodology.
- Staff turnover is manageable. Salary levels are competitive with other areas of public administration and with the private sector.
- Computing technology resources:
  - Sufficient resources are allocated to computing technology for hardware and software.
  - Each staff member has an assigned computer; additional computers are available for other users.
  - Back-up systems are in place to protect data.
  - Servers related to SIGFE are placed within the MOF building that is not adequately protected in the event of natural disasters, accidents, and other unusual events; authorities are taking corrective action to ensure business continuity.
- Office buildings, furniture, equipment, and transportation arrangements are adequate.
- Funding for identified needs of the statistical program is included in the overall budget of each compiling agency and is considered secure.

- Recommendations:
  - Allocate more human resources to compile and disseminate GFS.
  - Concentrate the responsibility to compile and disseminate GFS in one sector of the DIPRES, to avoid duplication of effort and improve efficiency.
  - Take corrective measures to ensure business continuity.

### 0.2.2 Measures to ensure efficient use of resources
- A significant level of manual data processing still exists; existing computing technology is not used to its full potential for compiling GFS.
- The DIPRES should improve the automatic derivation of GFS categories from SIGFE to minimize processing errors and improve efficiency.
- Priority automatic derivations recommended: expense by functions of government (COFOG), and data on stocks of financial assets and liabilities.
- External assistance and assessments:
  - July, 2003: IMF’s Fiscal Affairs Department provided an assessment of fiscal transparency practices.
  - May 2005: IMF’s Statistics Department (STA) conducted a government finance technical assistance mission; reviewed migration strategy for GFSM 2001 implementation, made recommendations, and presented a course on GFSM 2001.
- Recommendation: Improve the automatic derivation of GFS categories from SIGFE.

### 0.3 Relevance
- Current fiscal statistics are geared toward compliance with existing legislations and regulations, and international statistical standards.
- The DIPRES monitors relevance of GFS via informal contacts (e-mail inquiries, personal contacts, press conferences); occasional feedback is used to meet users’ needs.
- No user advisory groups exist.
- A Statistical National Commission (CNE) was established with representatives from the MOF, OCG, CBCH, NSI, and other nonpermanent public and private participants to promote documentation on methodologies, review methodologies, and identify emerging data requirements; the extent to which the commission has helped meet users’ needs is unclear.
- The DIPRES regularly participates in statistical meetings and seminars organized internally and abroad by international and regional organizations.
- Recommendation: Implement a structured and periodic process of consultation with policy departments and other principal users, including those in the private sector, to review the usefulness of existing fiscal data and to identify emerging data requirements.

### 0.4 Other quality management
- Management is sensitive to all dimensions of quality; training is provided to staff through regional programs.53
- Beginning in 2007, in the context of the public sector program of management improvement, the DIPRES will be audited using the ISO 9001:2000 international quality standards.
- Monitoring processes:
  - Processes are in place to check the quality of source data entered in system databases; results are relayed to managers.
  - SIGFE includes built-in checks and reconciliation procedures.
  - Public exposure of fiscal data provides a strong incentive to ensure quality.
- Planning trade-offs:
  - DIPRES recognizes trade-offs among quality considerations (accuracy and timeliness).
  - The financial programming sector changed the source data for GFS on financial assets from OCG’s accounting records to operational data from the Treasury due to legal timeliness requirements.54

### 1. Assurances of integrity — Professionalism, Transparency, Ethical standards
- Principal rules governing ethical conduct:
  - Law No. 18834 of 1989 (Administrative Statute)
  - Law No.19653 of 1999 (Law of Administrative Probity)
- Law No. 18834: principles of probity, confidentiality in legally restricted affairs, prompt reporting of irregularities.
- Law No.19653: requires blameless official conduct; President, ministers, sub-secretaries, and governors must submit a declaration of interests within 30 days after taking office.
- Recruitment: competitive process based on qualifications; promotion partly based on performance and seniority.
- Participation in international seminars and courses is supported.
- Choices of sources and techniques are based solely on data and processing requirements; DIPRES GFS are mainly based on internal data systems such as SIGFE; "neither further manipulation nor estimation or extrapolation techniques are needed."
- Dissemination decisions are based on legal obligations and international reporting commitments.
- GFS publications are released at press conferences by the Director of the DIPRES; a government official responsible for press and communication monitors media coverage and coordinates responses to misinterpretations.
- Terms and conditions for compiling and disseminating GFS are included in the Budget Law and are publicly available on the DIPRES’s website; the legislation referred in Section 0.1.1 is publicly available except for the Reserved Copper Law.
- No government officials outside the DIPRES have access to the data before public release; early access is limited to the Director of the DIPRES to comment at press conferences (this early access is not formally communicated to the public but is known).
- All DIPRES statistical products are clearly identified and DIPRES requests attribution when its statistics are used or reproduced.
- Advanced notice has been given to the public when major methodological or source data changes were introduced (major changes in 2003 related to adoption of GFSM 2001); explicative notes and methodologies were presented in press conferences and are available electronically.
- Ethical guidelines:
  - The MOF and DIPRES do not have their own comprehensive ethical guidelines; staff are guided by Law No. 18834 and Law No.19653.
  - Internal Bulletin No. 34 of December 2005 reminds DIPRES staff to adhere to ethical principles in the Law of Administrative Probity.

### 2. Methodological soundness — Concepts, definitions, and scope
- Concepts and definitions used to compile GFS are broadly consistent with the GFSM 2001 methodology.
- Chile has implemented GFSM 2001 in several reforms to enhance fiscal transparency and is a regional leader in implementing international standards for government finance statistics.
- October 2003: DIPRES included an official migration plan in Report on Government Finance, 2004 Draft Budget Law to fully adopt GFSM 2001.
- Progress:
  - First stage of the migration plan completed at the end of 2004: GFS improved by reclassifying existing data into the GFSM 2001 framework.
  - Report on Government Finance Statistics 1987–2003 (published October 2004) included accrual data for:
    - (1) unconsolidated GFS for general government and nonfinancial public enterprises,55
    - (2) financial assets and liabilities of the public sector, including the CBCH,
    - (3) a description of the methodology used,
    - (4) tables on institutional coverage,
    - (5) additional information on contingent liabilities and tax revenues.
  - Second stage of the migration plan calls for compilation and dissemination of GFS on a full accrual basis and on a cash basis, including an integrated balance sheet for the central government; this second stage was scheduled for completion in late-2006 but faced delays mainly due to lags in the implementation of SIGFE. SIGFE released two operational modules that will significantly improve information quality for GFS.
- Scope:
  - Chile is a unitary state with three independent powers.56
  - Budgetary central government comprises centralized units (ministries), constitutional state bodies (Congress, Judiciary, OCG, Office of the Public Prosecutor, Electoral Court), and decentralized agencies; universities and research institutes are part of central government but have their own budgets (autonomous nonprofit institutions).
  - Extrabudgetary central government includes: Oil Price Stabilization Fund; transactions related to the Reserved Copper Law; accrued interest on recognition bonds; funds in administration by the CBCH.
  - Local governments and public corporations, including Banco Estado and the CBCH, comprise the rest of the public sector.
  - GFS in Chile are focused on general government; DIPRES and OCG disseminate fiscal data with broader coverage.
- Institutional structure excerpt (central government A. Central government (1+2+3)):
  - 1. Budgetary central government: Approximately 190 government institutions, the OCG, Judiciary, Legislature, Ministries, and Presidency. Particularly, it includes the Pension Reserves Fund, and the Economic and Social Stabilization Fund.57
  - 2. Extrabudgetary central government: Oil Price Stabilization Fund, accrued interest on recognition bonds, Funds in Administration by the CBCH, Reserved Copper Law, and universities and research institutes.

*Source: _cr07318 - 0.2       Resources (IMF PDF).*

### 3. Social Security Funds

### 3. Social Security Funds

### Institutional coverage and scope
- Entities listed under social security funds include: Directorate of Social Security for Chilean Police, National Fund for Welfare Pensions, Social Security Fund for National Defense Employees, and Social Security Standardization Institute.
- Local governments: 341 municipalities.
- Public sector coverage definitions used in reporting:
  - A. (implicit) General government subsectors include budgetary central government, extrabudgetary central government, social security funds, and local governments.
  - D. Nonfinancial Public corporations: includes state-owned companies defined by law, and mixed companies where the government equity holdings represent more than 50 percent.
  - F. Financial public corporations: includes the state-owned public bank (Banco Estado) and the CBCH.
  - Consolidation hierarchy referenced: A+B = C (General Government Sector); C+D = E (Nonfinancial Public Sector); E+F = G (Public Sector).

### Main findings: institutional and transactional coverage
- The OCG collects and disseminates consolidated fiscal data for the public sector based on financial statements and balance sheets reported by all public entities, but these data do not conform with GFSM 2001 principles; assessment in this report is based on data compiled and disseminated by the DIPRES.
- The DIPRES disseminates GFS following GFSM 2001 principles for: (1) budgetary central government, (2) extrabudgetary central government, (3) social security funds, and (4) local governments; it also disseminates GFS for nonfinancial public corporations and additional CBCH information.
- Institutional coverage gaps:
  - Universities and ancillary units serving local governments (corporaciones municipales) are not included in GFS despite being controlled by the general government; GFSM 2001 methodology requires recording their activities with the controlling sector (budgetary central government and local governments).
  - Nonfinancial corporations owned by the Ministry of Defense are not covered in the nonfinancial public corporation subsector: FAMAE, ENAER, and ASMAR.
- Transactional coverage gaps and deviations from GFSM 2001:
  - Transactions of military entities regulated by the Reserved Copper Law are partially compiled: flow transactions are included but the stock of corresponding liabilities are not included in central government debt.
  - Three specific deviations identified:
    - Recapitalized interest in fiscal promissory notes with the CBCH (deuda subordinada) is not recorded as interest expense of the budgetary central government.
    - The stock of liabilities of the budgetary central government excludes interest accrued and not due on debt instruments, and excludes the stock of recognition bonds (although additional information on recognition bonds is disseminated separately).
    - GFS does not follow GFSM 2001 treatment of nonautonomous defined benefit schemes (Directorate of Social Security for Chilean Police, Social Security Fund for National Defense Employees): liabilities equal to net present value of expected future payments are not included as central government liabilities; imputed social contributions are not imputed as expense; household social contributions and benefit payments are treated as revenue and expense rather than financing transactions.

### Dissemination practices and public reports
- DIPRES annual reports:
  - Report on Government Finance Statistics (“Estadísticas de las Finanzas Públicas”):
    - Covers central government (budgetary and extrabudgetary), consolidated central government, local governments, nonfinancial public corporations, and the CBCH.
    - Published annually around the middle of the year. The most recent publication was disseminated in August 2006; the next publication is scheduled for July 2007.
    - Includes methodological section; statement of general government operations annually; monthly and quarterly consolidated central government operations; annual GFS functional classification of expense for consolidated central government; annual economic classification of tax revenue; annual data on financial assets and liabilities of consolidated central government and CBCH; human resources statistics; appendices on institutional coverage.
    - Published in hard copy and available on DIPRES’s website in PDF format; time series are available in Excel-readable format.
  - Report on Government Finance, Draft Budget Law (Informe de Finanzas Públicas, Proyecto de Ley de Presupuestos del Sector Público):
    - Disseminated annually in the last quarter of the year. The most recent publication was disseminated in October 2006.
    - Includes annual data on financial assets and liabilities of consolidated central government broken down by currency and instrument; detailed data and methodological notes on central government contingency liabilities.
    - Published in hard copy and available on DIPRES’s website in PDF format.
- DIPRES quarterly reports:
  - Statement of consolidated central government operations:
    - Disseminated within 30 days after the end of the reference period.
    - Contains: (1) revenue, expense, and transactions on nonfinancial and financial assets and liabilities following economic classification, broken down by budgetary and extrabudgetary and consolidated central government; (2) detailed tax revenue information; (3) stocks-and-flows data on Pension Reserves Fund; (4) stocks and flows on the Economic and Social Stabilization Fund and the Oil Price Stabilization Fund; (5) data on gross debt of the consolidated central government; (6) detailed budget execution by entities.
    - Broadly follows GFSM 2001 principles; available in PDF and time series in Excel-readable format.
  - Central government debt (Deuda del Gobierno Central):
    - Disseminated within 90 days after the end of the reference quarter.
    - Includes annual and quarterly data for budgetary central government gross debt, broken down by domestic and external debt, by currency, instrument, residual maturity, and by type of creditor.
    - Note: The breakdown of budgetary central government debt in domestic and external does not conform to the residence criteria: domestic and external debt correspond to debt issued domestically and abroad, respectively; amounts issued abroad and held by residents, and/or issued domestically and held by nonresidents, are not adjusted in GFS.
  - Report on public debt (Informe de Deuda Pública):
    - Disseminated within 90 days after the end of the reference quarter.
    - Includes data on gross and net debt of the budgetary central government and the CBCH by currency and instrument; gross and net debt of nonfinancial public corporations by maturity; amortization schedules for budgetary central government and the CBCH by currency; additional data on central government guaranteed debt to the rest of the public sector.
  - Report on financial assets of the Treasury (Informe de Activos Financieros del Tesoro Público):
    - Disseminated within 30 days after the end of the reference quarter.
    - Includes financial assets of the Treasury broken down by instrument, currency, and residual maturity; includes assets related to the Pension Reserves Fund and the Economic and Social Stabilization Fund.
  - Report on the Pension Reserves Fund and the Economic and Social Stabilization Fund (“Informe Ejecutivo Fondo de Reserva de Pensiones y Fondo de Estabilización Económica y Social”):
    - Disseminated within 30 days after the end of the reference quarter.
    - Includes detailed data on the financial assets of both funds.
- DIPRES monthly report:
  - Statement of budgetary central government operations (Estado de Operaciones del Gobierno Central Presupuestario):
    - Disseminated within 30 days after the end of the reference period.
    - Contains monthly data on revenue, expense, and transactions on nonfinancial and financial assets and liabilities presented according to GFSM 2001 classification; published in hard copy and PDF; time series available in Excel-readable format.
- Note: GFSM 2001 prescribes three basic statements: operations, integrated balance sheet, and sources and uses of cash. Chilean GFS include only the statement of operations for different public subsectors. An STA technical assistance mission in May 2005 prepared the three basic statements for consolidated central government for 2000–2004 and recommended that authorities verify and publish the results.

### Classification and sectorization issues
- Institutional sector definitions broadly follow 1993 SNA and subsectors follow GFSM 2001, with exceptions.
- Asymmetries between GFS and national accounts (NA):
  - Some entities sectorized in GFS within general government are treated as public corporations in NA: CORFO, Casa de Moneda, Direccion de Crédito Prendario, Central de Abastecimiento.
  - NA include in central government entities excluded from central government in GFS: Cajas de Compensación and Hospitales de la Ley de Accidentes de Trabajo.
- Recognition bonds sectorization discrepancy:
  - Recognition bonds are liabilities of budgetary central government to workers transferred from the former pension scheme to AFPs; related transactions should be recorded in budgetary central government, but interest accrued on these bonds is sectorized as extrabudgetary central government transactions, hampering subsector reconciliation.
- National classification practice deviations from GFSM 2001:
  - Revenue from copper (cobre bruto) is presented separately; it includes transfers from nonfinancial public copper enterprises comprising both taxes and dividends, which GFSM 2001 requires to be reclassified as taxes on income, profits and capital gains, and property income, respectively.
  - Expense data on subsidies and grants are aggregated in GFS; GFSM 2001 requires subsidies and grants to be reported separately with specified breakdowns (transfers to public/private enterprises; transfers to foreign governments/international organizations/other general government units; grants divided into current and capital).
  - Capital transfers are presented as net acquisition of nonfinancial assets; GFSM 2001 treats central government capital transfers to private sector and local governments for capital formation as expenses affecting the net operation balance, not as acquisitions of nonfinancial assets by general government.
  - Financing transactions and stock of financial assets and liabilities do not follow GFSM 2001 classifications; presented by residency and instrument summary form that does not conform to GFSM 2001.
  - GFS data on “reserved expenses” are not broken down into components due to lack of source detail; reserved expenses have been reduced significantly in recent years.
- COFOG (function of government) classification:
  - GFS follow international standard COFOG but are compiled and disseminated only for consolidated central government.
- Recommendations on classification/sectorization:
  - Apply consistently the sectorization of entities comprising the general government in all sets of macroeconomic statistics.
  - Follow GFSM 2001 guidelines in classifying revenue, expenses, and transactions on nonfinancial and financial assets and liabilities, both for stocks and flows.
  - Classify all transactions related to recognition bonds in the budgetary central government.
  - Compile and disseminate the classification of expenses by functions of government (COFOG) for local governments on an annual basis.

### Basis for recording and valuation
- Valuation:
  - Flows are recorded at market value, gross debt is recorded at face value, and financial assets are recorded at market value.
  - Debt denominated in foreign currency is converted into national currency using the average (between buying and selling) market exchange rates for the end of the period.
- Interest on indexed debt:
  - Recording of interest on indexed debt instruments does not conform to GFSM 2001 principles: GFSM 2001 requires that for index-linked debt the total amount paid or received corresponding to interest and any change in principal due to indexation be recorded as accrued interest. In Chile, accrued interest on instruments denominated in national currency and indexed to UFs, and debt instruments indexed to foreign currency, do not include changes in principal due to changes in the index.
- Recording basis:
  - Transactions of the public sector are recorded on an accrual basis in general; however, tax revenue is recorded on a cash basis, following a prudential criterion.
- Grossing/netting:
  - Transactions are shown on a gross basis, except refunds of taxes and overpaid expenses, which are shown as net, consistent with GFSM 2001; GFS for financing transactions are presented on a gross basis.
- Recommendations on basis for recording:
  - Follow GFSM 2001 principles in valuating stocks of assets and liabilities.
  - Record tax revenue on an accrual basis.
  - Record interest on indexed debt instruments according to GFSM 2001 principles.

*Source: _cr07318 - 3. Social Security Funds*

### 3. Accuracy and reliability

### 3. Accuracy and reliability

### 3.1 Source data
- Data sources are broadly adequate to allow the compilation of GFS for the general government and its subsectors.
- When a debt is denominated in a foreign currency, gains or losses resulting from exchange rate variations are recorded as other economic flows (OEFs). In this case, the amount of interest paid or received, which is recorded as a revenue or expense, does not include exchange rate variations and, consequently, leaves the operating balance and net lending/borrowing unchanged.
- Budgetary central government operations are recorded in SIGFE in accordance with the accounting principles and regulations set by the OCG, using a harmonized chart of accounts. SIGFE provides comprehensive accounting records, both to the DIPRES and the OCG, for compiling complete GFS, as well as budgetary execution reports on a monthly, quarterly, and annual basis.
- Although SIGFE reports reconciled data on flows and stocks for budgetary central government, the DIPRES compiles GFS on financial assets and liabilities position based on source data directly from the Treasury. By using nonharmonized data to compile stock statistics, the DIPRES is introducing a source for internal inconsistency between stocks and flows that is finally reflected in GFS.
- Data for extrabudgetary central government are collected quarterly by the DIPRES directly from the units that control the resources applicable to the corresponding fund or unit (the Treasury, the CBCH and the Social Security Standardization Institute), and from the Superior Council of National Defense for those transactions related to the Reserved Copper Law.
- For local governments annual source data are provided to the DIPRES by the OCG in an adequate and timely manner. Although there is no single chart of accounts applicable for the entire general government, charts of accounts used by all different levels of government are reconcilable and bridge tables have been constructed to compile consistent GFS.
- Data for nonfinancial public corporations are prepared under private sector accounting standards. Nonfinancial public corporations report to the OCG financial and balance sheet data in accordance with a harmonized chart of accounts; the quarterly report on public corporations is prepared by the Public Corporations System using the financial statements required by the Superintendency of Securities and Insurance. Data for each corporation, as well as consolidated data, are submitted to the DIPRES for review and then delivered to the Congress budgetary commission.
- The institutional and geographical coverage is broadly complete (with the exception noted in 2.3.1). A comprehensive register of individual public sector units is also available in the annual Budget Laws.
- The development of SIGFE has enhanced the integration of accounting records, budgetary data, and GFS. Although GFS codes are not currently embedded in the national chart of accounts, appropriate bridge tables have been constructed and included in SIGFE that facilitate the compilation of GFS directly from the underlying accounting data.
- A new chart of accounts was developed in 2005 as a coordinated effort between the budgetary and the statistical sectors of the DIPRES and submitted to OCG for review, which guarantees the consistency between accounting, budgetary, and GFS data.
- The DIPRES and OCG regularly review data sources for budgetary operations to ensure that SIGFE’s data collection systems remain comprehensive, but such policies and reviews are not documented. Maintenance of SIGFE, due to updates in the chart of accounts or changes in the accounting procedures and regulations, is done on a regular basis.

Recommendation (in text): Improve the valuation module of SIGFE to align current accounting data to GFSM 2001 principles. Stock should be valued at market prices.

### 3.1.2 Source data conformity and valuation
- Data sources for general government are consistent with guidelines on the definitions, scope, sectorization, classification, and time of recording.
- Charts of accounts for all levels of general government provide sufficient detail to align them with GFS classifications.
- Valuation of source data of stocks does not conform to the GFSM 2001 principles: SIGFE provides data on flows valued at market value, while stocks of financial assets and liabilities are recorded in the system at face value.
- The Treasury’s accounting records on financial assets and liabilities, used by the DIPRES to compile net debt statistics, are also at face value. The DIPRES uses complementary data to adjust the valuation of financial assets, recording them at market prices. However, liabilities are reported at face value.
- Source data are adequate to allow proper consolidation of different levels of government.

### 3.1.3 Timeliness of source data
- Source data are available with sufficient timeliness and periodicity to meet dissemination standards of GFS for general government and its subsectors.
- Timely data collection from all units comprising the central government and local governments is guaranteed by the CGO’s legal framework for reporting budgetary execution data.
- Data for budgetary central government are available almost in real time in SIGFE, reducing lags in publication of GFS and making the current flexibility option for the timeliness of central government operations nonbinding.
- SIGFE employs follow-up procedures to ensure timely delivery of source data (e.g., by maintaining regular contact with the reporting units).

### 3.2 Assessment of source data
- SIGFE uses Oracle software that incorporates extensive built-in cross checks to monitor data accuracy; validation procedures based on budgetary regulations are embedded in the system.
- Data inconsistencies and out-of-trend values are confirmed with the reporting units. Budgetary analysts check source data to further ensure accuracy. Most data are entered directly into the system and reports are generated automatically.
- Although source data for budgetary central government are carefully assessed through SIGFE, there are no audited records either from the OCG or from external audits. The auditing function of the OCG has not been fully developed and resources have been applied to this function on a selective basis.
- A Draft Law No. 522-354 of December 2006, currently being discussed by the Congress, proposes that the OCG is no longer responsible for the accounting of the public sector, and focuses on internal audit function.
- In 2005, the authorities, in collaboration with the Bank for International Settlements (BIS), undertook a reconciliation exercise for public external debt between national source data and creditor information that significantly improved the quality of fiscal data.

### 3.3 Statistical techniques
- Data compilation is mostly computerized, with built-in checks to minimize processing errors. Data collection is comprehensive for all subsectors and therefore there is no need to estimate missing data.
- Annual data are considered final when first released. Monthly and quarterly data are revised periodically and discrepancies or errors are corrected accordingly. However, compilation procedures are not documented.
- Other statistical procedures are rarely necessary. The DIPRES makes the following adjustments:
  - NA data for consumption of fixed capital provided by the CBCH are adjusted to account for coverage discrepancies with GFS.
  - Data on financial assets are adjusted to record stocks at market value.

### 3.4 Assessment and validation of intermediate data and outputs
- In SIGFE, procedures are in place to ensure that accounting data are accurate and final reporting of GFS is based on the most accurate and reliable source data.
- Large fluctuations or discrepancies are investigated. Expense by economic type and by functions, and transfers paid by central government to local governments are cross-checked for consistency.
- Discrepancies between subannual and annual data are identified and reconciled. However, discrepancies between flows and stocks are not monitored. The different valuation method used in source data for flows and stocks hinders a complete reconciliation.
- Fiscal data and other macroeconomic data, except for NA data of the CBCH, are not usually compared or reconciled.
- Data on selected external public debt stocks and related flows are compared with creditor information in a nonroutine manner.

Recommendation (in text): Routinely reconcile fiscal data with data in other macroeconomic statistical areas and periodically document status of reconciliations.

### 3.5 Revision studies
- Revision studies are not routinely conducted. Annual data are considered final when first published.
- Monthly and quarterly budget execution data and GFS are compiled from a comprehensive data collection system (SIGFE); therefore revisions are few and usually associated with corrections.

### 4 Serviceability

#### 4.1 Periodicity and timeliness
- Chile meets the SDDS requirements as follows:
  - Data on general government operations are disseminated annually.
  - Data on budgetary central government operations are disseminated monthly.
  - Data on central government gross debt are disseminated quarterly.
- Chile also disseminates quarterly data on consolidated central government operations, and net debt, as well as quarterly detailed data on financial assets of the Treasury.
- Timeliness meets SDDS requirements:
  - Annual data on general government operations are disseminated within two quarters after the end of the year.
  - Monthly data on budgetary central government operations are disseminated within one month.
  - Quarterly data on central government gross debt are disseminated within one quarter.
- In addition, Chile disseminates quarterly data on consolidated central government operations within one month after the end of the reference period and quarterly data on net debt within one quarter after the end of the reference period.
- Chile is availing itself of the flexibility option for timeliness of monthly central government operations.

#### 4.2 Consistency
- Monthly, quarterly, and annual GFS follow the same concepts, definitions, and classifications. Aggregates are consistent with their components, and balances are consistent with these aggregates.
- The sum of the transactions data for subsectors of the general government, after consolidation, is consistent with data on the consolidated general government.
- Vertical consistency: the statistical discrepancy between above-the-line (revenue and expense) and below-the-line data (net acquisition of nonfinancial assets and financing transactions) is not significant due to use of consistent source data.
- Horizontal consistency: consistency between debt stocks and financing is not held for the general government and its subsectors due to different data sources and valuation methods used to compile both sets of GFS.

Recommendation (in text): Reconcile periodically debt stocks and flows to improve internal horizontal consistency of GFS.

- Consistent time series are available from 1990 for annual data, from 1990 for quarterly data, and from 2003 for monthly data. For monthly data, consistent data exist but are not disseminated by the DIPRES in its publications.
- Changes arising from major methodological improvements are explained before the series are released. Migration to the GFSM 2001 methodology was broadly documented and disseminated to data users.
- GFS are reconcilable with NA data. In May 2005, an STA technical assistance mission helped the authorities with the reconciliation of NA aggregates with GFS categories for the period 2002—2004. The reconciliation results were published as an internal document of the CBCH but have not been updated and are not done periodically.
- GFS and external debt for the general government are not consistent due to discrepancies in the residence criteria. In GFS the classification of general government debt in domestic and external debt does not conform to the residence criteria used in the compilation of external statistics; the adjustment included in external statistics to account for residents’ holdings of government securities issued internationally is not incorporated in GFS.
- No reconciliation is done between GFS and monetary statistics.

Recommendations (in text):
- Reconcile GFS periodically with NA statistics.
- Introduce the residence criteria to classify debt in domestic and external.
- Regularly check consistency between GFS and monetary statistics.

#### 4.3 Revision policy and practice
- A revision policy has not been explicitly articulated. Data are considered final when published and there is no dissemination of preliminary data.
- Monthly and quarterly data are revised periodically, but revisions are rare and usually associated with corrections. When identified, corrections are included in the following annual, monthly, or quarterly report, respectively.
- No data are identified as preliminary. Revised data are not usually identified as such.
- The DIPRES has documented and disseminated to the public major revision studies related to methodological changes. Minor revisions are explained by press conference at the moment of release to the public.

### 5 Accessibility

#### 5.1 Data accessibility
- The GFS disseminated by the DIPRES are presented in a clear manner. Annual GFS on general government transactions are presented at current prices, constant prices, and percentage of GDP, with consistent time series for the last 10 years.
- Quarterly data for the consolidated central government operations include information on real change rates relative to an equal period of the previous year, and percentage of GDP. Monthly data for the budgetary central government include comparison of budgetary and execution data, as well as real change rates relative to an equal period of the previous year.
- Annual and quarterly reports include supplementary data on tax revenues by categories, and the stock of financial assets and liabilities by instrument, maturity, and currency.
- Time series are available, mostly in Excel-readable format, and commentaries on current-period developments are included in the publications and presented by the Director of the DIPRES at press conferences.
- The scope and classifications of GFS for general government broadly follow the GFSM 2001 principles, with some significant exceptions. The level of detail by instrument provided in financing transactions and gross debt is less than recommended by international guidelines.
- DIPRES’s publications are available in hard copy as well as in electronic format on the official website. Data associated with the SDDS are disseminated on the DSBB.
- The DIPRES has an advance calendar for data dissemination posted on the official website at the beginning of the year. GFS are released punctually, according to a regular and preannounced schedule.
- GFS are made available simultaneously to all interested users. The public is informed of the statistics being released, and of the procedures to access them.
- GFS not routinely disseminated are made available upon request. Customized tabulations are not usually provided. The availability of additional statistics and the procedures for obtaining them are not made known.

*Source: _cr07318 - 3. Accuracy and reliability*

### 5.2       Metadata       accessibility

### 5.2       Metadata       accessibility

### Documentation and metadata availability
- The annual Report on Government Finance Statistics (Estadísticas de las Finanzas Públicas) includes extensive information on source data and methodology used in the compilation of GFS, with a methodological section regularly updated that contains information on concepts, definitions, classifications, data sources, compilation methods, and statistical techniques used to compile GFS.
- Other publications contain comprehensive descriptions of sources and methods of more specific fiscal data.
- The SDDS metadata for GFS were updated on March 2007.
- The names of the contact persons for central and general government operations, and central government debt are accurate.
- Metadata on methodologies need to be updated.

### Levels of detail and target audience
- Only the information detailed in 5.2.1. is available.

### Assistance to users — current practices and recommendations
- No contact person for GFS is identified in the DIPRES’s publications.
- Contact persons and contact details for each subject field are publicized in the SDDS metadata on the IMF’s DSBB.
- In general, broad assistance is provided to general public users of GFS.
- Recommendation: Contact persons and contact details should be included in the DIPRES’s publications.
- There are no catalogs of publications, documents, and other services.
- Most statistical publications are available free in hard copy and on the DIPRES’s website.
- Recommendation: Publish a catalog of publications.

### Data Quality Assessment — summary findings (Table 9)
- Prerequisites of quality
  - Legal and institutional environment: “There is room for improvement in data sharing and coordination among the various units involved in the compilation process.”
  - Resources: “At present, the number of staff devoted to produce GFS is barely adequate.”
  - Relevance: “There are no user advisory groups.”
  - Other quality management: noted.
- Assurances of integrity: Professionalism, Transparency, Ethical standards — observed/labeled.
- Methodological soundness
  - Concepts and definitions: observed/labeled.
  - Scope: “Institutional and transactional coverage for the general government are incomplete.”
  - Classification/sectorization: “Sectorization is broadly in line with the GFSM 2001. However, significant deviations exist in national classification practices.”
  - Basis for recording: “Tax revenues are recorded on a cash basis. Interest of indexed debt instruments do not include changes in principal due to fluctuations of the index.”
- Accuracy and reliability
  - Source data: “Stocks of financial assets and liabilities are recorded at face value.”
  - Assessment of source data and Statistical techniques: noted.
  - Assessment and validation of intermediate data and statistical outputs: “The process for validating data is not well established.”
  - Revision studies: noted.
- Serviceability
  - Periodicity and timeliness: noted.
  - Consistency: “GFS are not internally consistent due to differences in source data used to compile debt flows and stocks.”
  - Revision policy and practice: noted.
- Accessibility
  - Data accessibility, Metadata accessibility, Assistance to users: noted.
  - “Catalogs of publications are not available to users of statistics.”

*Source: _cr07318 - 5.2       Metadata       accessibility*

### 0.3       Relevance

### _cr07318 - 0.3       Relevance

### Relevance of statistics and user engagement
- The CBCH does not have a policy of periodically consulting users, but users can send comments and suggestions through the CBCH’s website via a specific window.
- CBCH staff regularly participates in national and international seminars and courses, including those offered by the IMF.
- Data collected by the SBFI are mainly for supervisory rather than for statistical purposes, making the SBFI the main user of the data it collects.
- The SBFI has adapted its requirements from ODCs to meet the CBCH’s emerging needs on monetary statistics.
- The SBFI publishes financial data on its website and provides users with analytical tools; SBFI staff regularly attends technical meetings on banking supervision.

### Other quality management
- The CBCH produced a strategic plan for the period 2006–2008; one objective: “to contribute to an understanding of economic and financial affairs by preparing and broadcasting relevant reports, studies and statistics.”
- The Statistical Information and Research Unit’s strategic objective: “to compile and publish national, regional, and sectoral statistics pertinent to the scope of the CBCH and to the conduct of its monetary policy, maintaining high standards of quality, timely information and using the best technology available.”
- GEMF staff are aware of data quality dimensions; CBCH publishes data as provisional during two months and after that presents them as final.
- The GIIE reports systematically to CBCH authorities any developments that may affect statistical quality and requests outside assistance when needed (for example, IMF consultations during the recent revision of the monetary aggregates).
- The GIIE working program outlines developmental and functional objectives; one ongoing project focuses on statistics quality.
- GEMF systematically checks aspects such as coverage to guarantee relevance.
- User comments are incorporated via ad-hoc projects; example: GIIE prepared a document and a new set of tables with the CBCH’s balance sheet information in response to user requests and distributed a preliminary version for comments before public dissemination.

### Assurances of integrity — Professionalism
- By Constitution (Chapter XIII, Article 108), the CBCH is an autonomous entity; the SBFI is an autonomous institution with financing from supervised institutions—legal provisions guarantee independence of both institutions.
- Technical criteria alone are applied by data compilers and analysts; both institutions are independent in choice of data collection methods.
- Staff of both institutions is hired on a permanent basis following a competitive process, with openings posted on their websites; internal promotions follow a competitive process considering professional qualifications and past performance.
- CBCH training policy: staff participates in internal seminars, national and international courses, and is supported to pursue postgraduate studies abroad; emphasis has shifted toward specific courses and seminars. Staff is encouraged to publish articles.
- Source data for monetary statistics: balance sheets of the CBCH, the 25 commercial banks, the state owned Banco Estado, the five credit unions, and the 21 mutual funds; full coverage restricts manipulation through sample changes.
- Non-accounting financial statistics (e.g., interest rates) are produced using responses by all institutions of the system.
- GEMF follows the methodology of the Monetary and Financial Statistics Manual (MFSM) for classification and sectorization.
- Dissemination follows established procedures and an advance release calendar; periodicity and timing were never hampered by pressures from other government agencies.
- In case of misinterpretation or misuse, CBCH provides comments and clarifications mainly through its Communications Management Office (GC); GC follows daily press and provides summaries electronically to staff.

### Assurances of integrity — Transparency
- Chile participates in the IMF’s Special Data Dissemination Standard (SDDS); metadata on monetary statistics are accessible on the IMF’s DSBB and were last updated in December 2006. CBCH’s website links to the country page on the DSBB.
- Legal frameworks: Law 18840/89 and DFL 3/97 specify legal terms and are posted on CBCH’s and SBFI’s websites. Agreement 1110-02-040122 and successive modifications specifying monetary statistics content, periodicity, and timeliness are posted on the CBCH’s website.
- The SBFI’s website contains legal regulations and a description of departmental responsibilities.
- There is no prior governmental access to published monetary data. GEMF prepares high-frequency monetary indicators distributed to CBCH authorities and the Minister of Finance; some are not published and are only estimates.
- Article 14 of DFL 3/97 allows the SBFI to provide information to the Minister of Finance and the CBCH notwithstanding banking secrecy, intended for institutions facing solvency problems and does not affect dissemination of monetary statistics.
- CBCH and SBFI publications are clearly identified with logos and addresses. Tables in the CBCH’s Monthly Bulletin show the data source when external to CBCH but do not identify the internal unit responsible (identified in an explanatory note).
- CBCH informs methodological changes through meetings and seminars with the press and analysts; interested parties receive an advance copy for comments before public release. Major methodological changes are published in Estudios Económicos Estadísticos.

### Assurances of integrity — Ethical standards
- CBCH’s Personnel By-Laws detail obligations and prohibitions, including confidential handling of information, restrictions on electronic mail use, sexual harassment, and other working arrangements.
- Serious breaches include substance abuse, alteration of accounting records, and noncompliance with personal financial obligations.
- CBCH staff cannot have other paid employment except academic positions; conflict of interest rules extend to spouses—spouses are forbidden from entering into commercial relationships with CBCH.
- Sanctions can include termination. The CBCH’s Ethics Code deals with incompatibilities, conflict of interest, safeguarding information, communications with the press, participation in seminars, use of working elements, and gifts.
- New employees receive an induction on these restrictions.
- A 2003 breach of confidentiality (privileged information transmitted to private investors by an assistant to the President) led to criminal prosecution of the assistant and the President’s resignation; controls were subsequently heightened.

### Methodological soundness — Concepts and definitions
- GEMF methodological framework for central bank and ODCs surveys is consistent with MFSM guidelines.
- A depository corporations survey is compiled consolidating the central bank survey with the ODCs survey.
- Main monetary aggregates, revised in mid-2006:
  - Monetary base: notes and coins issued by the central bank plus deposits of ODCs with the central bank.
  - M1: notes and coins in circulation outside the banking system, checks issued by the central bank, demand deposits of the private sector in national currency (net of checks to be cleared), other sight deposits of the private sector in national currency, and savings deposits of the private sector in national currency with a maturity of less than 30 days.
  - M2: M1 plus time deposits of the private sector in national currency, savings deposits of the private sector in national currency with a maturity longer than 30 days, mutual funds’ shares of the private sector with a maturity of less than one year, plus credit unions’ deposit liabilities, less investment of mutual funds and credit unions in instruments included in M2.
  - M3: M2 plus foreign currency deposits of the private sector, central bank bills held by the private sector in other depository corporations, treasury promissory notes held by the private sector in other depository corporations, letters of credit held by the private sector in other depository corporations, mutual funds’ shares of the private sector with a maturity longer than one year, plus voluntary savings quotas of pension funds, less investment of mutual funds and pension funds in instruments included in M3.

### Methodological soundness — Scope
- At end-April 2007, Chile’s depository corporations sector comprised the CBCH, 25 commercial banks, the state owned Banco Estado, five credit unions, and 21 mutual funds.
- There are no banks in liquidation. Coverage of the ODCs survey is complete, comprising the headquarters and all domestic branches.
- Recommendation: Start compiling an OFCs survey, with data from pension funds, insurance corporations, and other relevant financial intermediaries.
- Current practice: GEMF compiles a survey of the pension funds (using data from the Superintendency of Pension Funds—SPF) in line with MFSM recommendation; this should be expanded to include insurance corporations and other financial intermediaries, with data collected by the Superintendency of Securities and Insurance (SSI).
- The SSI transmits data on insurance corporations, health insurance companies, and investment funds, but in a format not adequate for compiling an OFCs survey; CBCH should request SSI to review the format.

### Methodological soundness — Classification / sectorization
- Sectorization follows MFSM and BPM5 recommendations. Identification of residents/nonresidents and resident sectors is facilitated by the uniform tax register (Registro Único Tributario—RUT).
- GEMF supplements monthly balance sheet data with additional disaggregation by instrument and sector to sectorize ODCs accounts.
- CBCH accounting records do not detail holders of securities issued by the CBCH; sectorization is done by subtracting holdings reported by ODCs from total issued by CBCH and assigning the difference to OFCs.
- Ongoing project: improve sectorization of CBCH-issued securities by identifying final holders and grouping them into commercial banks, pension funds, investment funds, insurance corporations, and foreign corporations; holdings by national nonfinancial corporations calculated by difference.
- Shortcoming: nonsectorization of mutual funds’ liabilities, lumped into one item Participations in mutual funds. In the ODCs survey, these liabilities are classified as securities held by OFCs.
- Recommendation: Properly sectorize the securities issued by the CBCH according to their final holder, and the liabilities of mutual funds.
- Classification of financial instruments follows MFSM recommendations; instruments classified by type, maturity, and currency (local currency, indexed to UF, and foreign currency).
- Repurchase agreements classified as collateralized loans; financial derivatives properly recorded; advances on insurance premiums recorded as Other assets rather than Insurance technical reserves.
- Valuation adjustments are transferred to profits and loss accounts.
- Proportion of Other assets and Other liabilities: less than 0.5 percent of CBCH’s balance sheet; around 7.5 percent of aggregated ODCs balance sheet.

### Methodological soundness — Basis for recording
- Market prices are generally used to value flows and stocks.
- Accounts in foreign currency are converted into national currency at the end of the month using the CBCH’s “observed” peso per U.S. dollar exchange rate, calculated as the weighted average of interbank transactions (excludes operations with the CBCH), resulting in a mid-point market exchange rate.
- ODCs value their holdings of securities at market prices, consistent with the MFSM.
- For prudential reasons, the CBCH values its holdings of securities at the lower of the market price or the amortized cost (acquisition cost plus accrued interest). Securities issued by the CBCH or ODCs are valued on the liability side at amortized cost.
- Gold valuation: CBCH values monetary gold at the end of the month on the basis of the simple average of the international gold prices of the last 90 days, then discounted by 10 percent and converted into national currency using the “observed” exchange rate of that day. This is a conservative approach not fully in line with MFSM (which recommends market prices); the difference is nonmaterial due to the small stock of monetary gold held by the CBCH (0.02 percent of the CBCH’s international reserves).
- Recording is on an accrual basis; accrued interest is presented with its underlying instrument. For prudential reasons, interest is not accrued on loans overdue for more than 90 days.
- Grossing/netting: assets and liabilities of CBCH and ODCs are presented on a gross basis with no netting of claims against liabilities to the same transactor or group. Accounts in overdraft are recorded and reported as loans. Loans are recorded gross at nominal values plus accrued interest, with provisions for loan losses reported as a separate item under Other liabilities.
- In the ODCs survey, GEMF consolidates intrasectoral accounts of commercial banks, credit unions, and mutual funds in Other items net. The DC survey consolidates accounts of the CBCH with ODCs, also in Other items net.

*Source: _cr07318 - 0.3       Relevance (Source PDF: _cr07318 - 0.3       Relevance)*

### 3. Accuracy and reliability

### 3. Accuracy and reliability

### 3.1 Source data
- Source data are obtained from comprehensive data collection programs that take into account country-specific conditions.
- Use of a common chart of accounts and the RUT simplifies compilation of monetary statistics.
- The SBFI has issued precise instructions on the different forms ODCs must complete, the use and recording of the accounts, validating procedures, and timeliness.
- Codification of accounts allows classification by instrument, sector, maturity, currency denomination, type of operation, risk, etc.
- To prepare the analytical accounts of the CBCH, the GEMF uses an automated program (ANABAL) that reorders trial-balance accounts according to user needs.

Findings on data recording and standards:
- The depository corporations survey is based on actual accounting records of the CBCH and the ODCs; no sampling or estimations are employed by the GEMF.
- Supplemental information with monthly balance sheets permits adequate sectorization and identification of residents and nonresidents.
- Reporting errors are minimized by use of the RUT of the account holder.
- ODCs follow national accepted accounting standards and are in the process of adopting International Financial Reporting Standards (IFRS); the migration to IFRS would imply minimal changes.
- The CBCH balance sheet is prepared in accordance with national accepted accounting standards and audited by an external international firm selected through a competitive process for a period of three years, renewable on a yearly basis for up to two years.
- The 2006 financial statements were approved with an unqualified opinion.

Timeliness of source data:
- The CBCH’s data are available on a daily basis to the GEMF, which has online access to accounting records.
- ODCs transmit daily data to the SBFI—with a one-day lag—which are utilized by the GEMF for some indicators.
- ODCs’ monthly balance sheets are transmitted within eight working days to the SBFI, and forwarded also with a time lag of eight working days to the CBCH.
- The whole process is automated, allowing the GEMF to produce timely monetary data, fully in compliance with SDDS requirements.

### 3.2 Assessment of source data
- Before submitting data to the SBFI, ODCs run internal consistency checks with programs provided by the SBFI.
- After receipt, the SBFI performs automatic validation and consistency checks: comparison of assets and liabilities, logical validation, cross-checking of interbank positions, and data analysis.
- The GEMF runs automatic consistency checks at the level of individual institutions, with macros flagging discrepancies; GEMF contacts reporting institutions to resolve inconsistencies.
- Daily data reported by ODCs are analyzed to detect unusual movements and used to verify accuracy of monthly data.

### 3.3 Statistical techniques
- Both CBCH and ODCs data are processed electronically, minimizing the risk of mistakes.
- Monthly data reported by ODCs are complete; the GEMF does not estimate missing information.
- For missing daily data from ODCs, the GEMF delays (by hours) posting indicators and waits for data if they pertain to a large institution, or repeats the value of the previous day for small institutions.
- Because the depository corporations survey is complete and balance-sheet based, there is no need for data adjustments or transformations.
- Monetary data are seasonally adjusted using the X-12-ARIMA procedures.

### 3.4 Assessment and validation of intermediate data and statistical outputs
Validation practices and gaps:
- Interbank positions are validated by the SBFI, but not at an aggregated level by the GEMF.
- Data reported by ODCs on deposits with the CBCH are not cross-checked against the CBCH’s own accounting records.
- Data reported by ODCs on securities issued by the CBCH cannot currently be cross-checked against CBCH data because the latter do not exist; GEMF must accept ODC-reported data.
- The GEMF checks daily that changes in international reserves (calculated by DBPDE) are consistent with monetary data; some discrepancies arise due to different valuation methodologies, but they are explained.
- Consistency checks between monetary and fiscal data are performed by the Macroeconomic Analysis Unit (UAM), not by the GEMF. The UAM controls MOF data on debt to and deposits with the CBCH and contacts the MOF to resolve discrepancies.
- The UAM does not cross-check monetary data on loans to government by ODCs, central government securities held by ODCs, and central government deposits with ODCs against fiscal data.

Recommendation (text preserved):
- Perform systematic validations of reported monetary data, including aggregated interbank positions, data reported by ODCs vis-à-vis the CBCH against its own accounting records, data reported by ODCs on holdings of central government securities against CBCH’s records on central government debt, and monetary and fiscal data on credit to government by ODCs.

Investigation and follow-up:
- If discrepancies arise during validation of ODC data, the GEMF contacts the reporting institution to solve the problem.
- Procedures exist to investigate classification/sectorization errors or omissions; if errors are detected, all involved parties are informed.

### 3.5 Revision studies
- Revision studies are not routinely performed owing to the nature of monetary statistics.
- The CBCH conducted some revisions in the recent past and informed users beforehand.
  - End-2006: new monetary tables introduced and some reformatted to better monitor foreign exchange interventions, international reserves, foreign currency exposure, and consolidated foreign liquidity position with the central government. The proposal was posted on the Internet for comments, explained at a press conference, discussed with experts, and a detailed explanation was published in Estudios Económicos Estadísticos at end-2006.
  - Mid-2006: a similar process occurred for revision of monetary aggregates.
- When new monetary aggregates were introduced in mid-2006, series were revised from January 1986 onwards. Series discontinued from publication are nevertheless updated internally to preserve historical continuity.

### 4. Serviceability

Periodicity and timeliness:
- Daily data on some financial indicators (including interest rates) are posted on the CBCH’s website.
- Monetary aggregates are disseminated twice a month by the CBCH: the first as an advance to the monthly data, and the second as the full release.
- Monetary aggregates are disseminated on the 7th and 23rd day of each month; the first release does not contain ODC data from the previous month and is intended as an advance. On the 23rd, the CBCH’s Boletín Mensual (containing monetary data) is posted on the website and later distributed as a hard copy.
- This periodicity and timeliness comply with SDDS requirements.
- More detailed monetary data are published once a year in the CBCH’s Síntesis Monetaria y Financiera.

Consistency:
- ODCs' claims on the CBCH are fully consistent with liabilities to ODCs recorded by the CBCH, but loans to ODCs recorded by the CBCH are considerably lower (on average only one fifth) than the liability to the CBCH for this concept reported by ODCs.
- A large discrepancy exists between claims on ODCs reported by ODCs and liabilities to ODCs reported by them, largely explained by nonsectorization of mutual funds’ liabilities.

Recommendation (text preserved):
- Systematically review aggregated source data reported by ODCs to eliminate internal inconsistencies.

Revisions policy and practice:
- Monetary data posted on the CBCH’s website and published in its Boletín Mensual are provisional for the last two months and made final after this period.
- Preliminary data are identified through a footnote in the tables.
- Studies and analyses of revisions are made public in the CBCH’s Estudios Económicos Estadísticos and explained through notes in the corresponding tables.

### 5. Accessibility

Data accessibility:
- Monetary statistics are first disseminated on the CBCH’s website and then published as a hard copy in the CBCH’s Boletín Mensual; more disaggregated data are published annually in the CBCH’s Síntesis Monetaria y Financiera.
- Tables present information clearly with an adequate level of detail; notes accompany all tables, including explanations due to revisions.
- The Boletín Mensual and its fortnightly advance are available through the CBCH’s website; Boletín Mensual and Síntesis Monetaria y Financiera are available in hard copy.
- Time series posted on the website are accessible in an Excel format and graphs can be generated.
- The latest available series at the time of the mission’s visit to Santiago correspond to March 2007.
- The CBCH posts a release calendar, which is strictly followed and included in its Boletín Mensual.
- Monetary statistics are released through the CBCH’s website, making them available to all users at the same time; the press is not informed in advance.
- Additional data may be made available upon request with authorization of the general manager.

Metadata accessibility:
- Metadata on monetary statistics are available on the IMF’s DSBB. The latest update of the metadata on the DSBB took place in December 2006.
- The CBCH’s website contains a link to the country page on the DSBB.
- The level of disaggregation in CBCH publications is sufficient to meet different audiences’ needs.

Assistance to users:
- Contact points for each subject field are available on the DSBB, accessible through the CBCH’s website.
- A catalog with the publications of the CBCH and their prices is posted on the CBCH’s website and included in its Boletín Mensual. Several publications can be downloaded from the Internet.

*Source: _cr07318 - 3. Accuracy and reliability*

### 0.1 Legal and institutional environment

### 0.1 Legal and institutional environment

### Responsibility for compiling and disseminating statistics
- Chile’s balance of payments are compiled and disseminated by the BCCH.
- Constitutional Organic Law of the BCCH (Law No. 18840 published in the official gazette on October 10, 1989) — Article 53 stipulates the Bank shall compile and publish, in a timely manner, the country’s principal macroeconomic statistics, including monetary, foreign exchange, and balance of payments statistics, and the national accounts.
- Coverage, periodicity, and timeliness of these statistics are established by Decision No. 393-07-941207 of the Central Bank Board of Directors, published in the official gazette on December 13, 1994, and last updated in 2006.
- The unit compiling the balance of payments within the BCCH is the Balance of Payments and External Debt Department (DBPDE), part of the Information and Statistical Research Directorate (GIIE), which reports to General Management.

### Data sharing and coordination among data-producing agencies
- Major external data provider: National Customs Service (NCS) — provides data on exports and imports; flow of source data regulated by an agreement between NCS and BCCH.
- Customs data handling:
  - Received as individual operations; processed by the Database Department (DBD) and the International Exchanges and Statistical Compilation Directorate (DCIE).
  - Entered into institutional database and retrieved by DBPDE for balance of payments and by National Accounts Department.
  - Customs documents transmitted via a dedicated telephone line ensuring weekly receipt.
  - Data checked for consistency prior to release; errors corrected with agreement of NCS.
  - Customs declarations systematically processed enabling use for flows of goods and for estimating/validating transport, export charges, and trade credits.
- Tourism data:
  - National Tourism Service (SERNATUR) agreement with BCCH since 1991; format and requirements amended to meet BCCH needs.
  - Historically gathered in three survey periods (high, medium, and low season) with findings delivered within a maximum of six months.
  - Since January 1, 2007 SERNATUR has been implementing the new methodology requiring delivery of findings monthly, quarterly, and with a three-month lag; work is on final step to amend to require year-round information gathering.
  - Supplemented by International Police data on number of travelers entering and leaving the country with geographical breakdown; compiled monthly and received with a lag of less than one month.
- Other external sources:
  - Superintendency of Pension Fund Administrators (SPF): monthly international portfolio buying and selling and a quarterly statement of changes in foreign investments by pension funds (balances, flows, other changes).
  - Superintendency of Securities and Insurance (SSI): external assets position of supervised entities and public data on income statements.
  - Budget Directorate of the Ministry of Finance (DIPRES), banking system: data received via SINACOFI or directly from banks; includes external assets and liabilities positions broken down by original maturities and categories (cash, deposits, loans).
  - Internal Taxation Service (ITS): no formal agreement; provides data on a case by case basis (e.g., data on additional tax used to estimate imports of various services items).
  - Foreign Investment Committee (CIE): provides data in response to official contacts to supplement other channels and establish breakdowns by country or sector.
- Centralized compilation and surveys:
  - International Exchange and Surveys Department (DCIE) compiles and validates external source data, managed by Compendium of International Exchange Regulations (CIER); processes surveys conducted by BCCH; reporters may consult and amend their previously provided information.
- Coordination and access:
  - Ongoing meetings between SERNATUR and BCCH and with other external providers reporting via the CIER.
  - Access to customs records is expeditious and without confidentiality restrictions; tax records available on an anonymous basis.
  - Quality and coverage of administrative records is good; appropriate filters developed for balance of payments compilations.
- Recommendation:
  - Provide more timely access to tax records to improve quality of quarterly estimates; DIPRES must take into account the statistical information requirements of the CBCH (balance of payments, international investment position, external debt, and national accounts) within the framework of Article 19 of the Fiscal Responsibility Law.

### Confidentiality of reporters’ data and internal controls
- Confidentiality guaranteed under Article 66 of the Organic Law of the CBCH.
- Manual of Ethical Principles and Personnel Regulations prohibit officials from divulging confidential information or information pending publication.
- Individual employment contracts require maintaining confidentiality; violation constitutes reasonable grounds for termination of employment.
- Survey communications inform respondents that information will be used for statistical compilation purposes and will remain confidential.
- BCCH data processing platform employs a hierarchical system of privileges to protect access to information sources and compilation processes; policies exist on “clean desks” and mechanical destruction of hard copy information.

### Legal mandate and measures to encourage response
- Legal authority:
  - Article 53 of the Organic Law authorizes the Central Bank to obtain information from public institutions for compiling the balance of payments; Article 40 authorizes the CBCH to request information on foreign exchange operations using the CIER.
- Private sector response and coverage:
  - No general legal means to require private institutions to respond for statistical purposes beyond foreign exchange operations; nevertheless, broad cooperation exists.
  - Response rates to DCIE surveys (qualitative response rate with quantitative response rate in parentheses):
    - portfolio investment 78 percent (95 percent);
    - direct investment 70 percent (70 percent);
    - international institutions excluding embassies 50 percent (100 percent);
    - services 64 percent (57 percent).
  - Given relative importance, improving receptiveness to direct investment and services surveys would enhance recording of nonfinancial private sector operations.
- Incentives and procedures to improve response:
  - Consideration of feedback mechanisms to reporters (e.g., daily bulletin published on website).
  - Preference for persuasion over sanctions for recalcitrant cases, even when sanctions exist for the CIER.
  - Recommendation: Seek to improve the legal framework authorizing the CBCH to request information for statistical purposes from the private sector as a whole to complement existing legislation applicable to the public sector.
  - CBCH designs forms and surveys to align with enterprises’ record-keeping to minimize respondent burden; some formats include a question on time to complete.
  - Pilot tests and mass meetings with reporters are used to ensure understanding (example: six meetings with approximately 400 people from major importing and exporting enterprises to explain new forms, Chapters IV and V of the CIER).
  - DCIE provides ongoing availability (telephone and e-mail) during response periods; thematic groups between DCIE and DBPDE meet periodically and are supervised monthly by management.
  - CBCH seeks alternative data sources (publicly available information from banks, Pension Fund Administrators, insurance enterprises) to avoid unnecessary demands on reporters.

*Source: 0.1 Legal and institutional environment*

### 0.2       Resources

### 0.2       Resources

### 0.2.1 Staff, facilities, computing resources, and financing are commensurate with statistical programs

- Organizational structure
  - DBPDE: two working groups—(1) compiling goods, services and transfer accounts; (2) income, financial, international investment position and external debt accounts.
  - DBPDE works closely with DCIE and DBD for source data collection.

- DBPDE staffing and qualifications
  - 20 full-time and two fixed-term professional staff.
  - 60 percent of professional staff have 10 or more years of experience in compiling external accounts statistics; the remainder have an average of less than five years experience.
  - Three very experienced administrative assistants support compilation.
  - Professional staff composition: commercial engineers (15), statisticians (3), industrial engineers (1), engineers in international trade (1), accountants/auditors (1), systems analysts (1).
  - More than one third hold post-graduate degrees (masters and doctoral candidates).

- DBPDE training and capacity building
  - Basic training for new professional staff; attendance at courses on balance of payments and external debt methodology (IMF and CEMLA).
  - Visits to centers of excellence: Office for National Statistical in England, Bank of Spain.
  - Participation in seminars (OECD and others); seven-month in-service training at the Bank of Spain.
  - Additional training: seminars and programs in Chile; one staff member enrolled in a Statistics Diploma program.
  - DBPDE organized a course on balance of payments methodology (November 2005) and a workshop on external trade price indices delivered by an international expert.

- CIER / DCIE staffing and qualifications
  - Post-2001 restructuring: former capital control units eliminated; CIER refocused on statistics.
  - DCIE: one fixed-term and 13 full-time and one fixed-time professional staff members.
  - Three very experienced administrative staff support data compilation.
  - DCIE professional staff composition: commercial engineers (9); accountants/auditors (3); civil engineers (1); statisticians (1).
  - More than one-third of DCIE professional staff have post-graduate qualifications; another four are currently pursuing post-graduate degrees in finance.
  - Consideration being given to requiring good working knowledge of English for potential new recruits.

- DCIE training history
  - Staff attended courses on financial markets, balance of payments statistics; monetary and financial statistics; processing of surveys, and sample design at IMF, CEMLA, and the National Institute of Statistics of Spain.
  - One-year in-service training programs in 2005 in the Netherlands, Sweden, and Denmark.
  - In-service training programs in 2002 in Australia, New Zealand, Spain, and Portugal.
  - Administrative staff continually upgrade technological and basic statistics skills.
  - Arrangements for a workshop with Head of Balance of Payments Department of the Bank of Spain and an internship on surveys at National Institute of Statistics of Spain.

- DBD staffing and role
  - DBD: 10 professional staff (commercial, industrial, and computer engineers) and two administrative staff.
  - Manages operational databases for external trade in goods statistics; coordinates relations with providers such as NSI, NCS, and SERNATUR.

- Compensation and human resources policies
  - CBCH compensation levels and structure competitive with market when including nonmonetary benefits (access to overseas training).
  - Mid-2006: all staff reclassified to make salary mobility more flexible.
  - Salary scale being adapted to achievement of objectives and to provide performance bonuses.
  - Staff assessed annually on teamwork, individual contribution, and achievement of previously defined objectives; recent revision aligned with salary mobility and performance bonuses.

- Computing resources and analytical tools
  - Individual staff desktop computers: at least a Pentium 4HT, networked Windows XP Operating Systems.
  - Peripherals: color printers and photocopiers; black and white printers and copiers for heavy-duty work; portable computers; overhead projectors.
  - Software: Microsoft Office (Excel, Word, Power Point, Access); EVIEWS, GAUSS, SPSS.
  - Corporate database tool: Sybase.
  - Institutional systems: Debt Management and Financial Analysis System (DMFAS); OLAP tools.
  - Projects underway to improve warehousing and calculation of standard balance of payments outputs and to incorporate FAME as institutional time series database.

- DBPDE Balance of Payments Database project
  - Purpose: repository of disaggregated balance of payments series (but not microdata) to automatically generate aggregated outputs and handle versions (historical, published, in progress, simulations).
  - Initial rollout planned in 2007-2008.
  - Platform: FAME Software to store and manage time series centrally for internal use and future public dissemination.
  - Expected benefits: efficiency gains, closer interaction between statistical units and IT staff, greater transparency, specialized statistical chain functions, facilitated analyst work.

- Data access, microdata, and compilation security
  - All staff connected to the Internet; active Intranet.
  - Need identified: development of efficient systems for storing and accessing microdata underlying aggregated series and for creating algorithms to perform calculations across basic information from various sources.
  - Recommendation: Design and create a data warehouse for external statistics (balance of payments, IIP, external debt) to replace numerous spreadsheets with complicated and unstable links and provide a reliable repository for large amounts of information.

- Physical facilities and budgets
  - Individual workstations meet international standards in size; adequate lighting, cooling, and heating.
  - Office furniture and equipment (desks, chairs, filing cabinets, telephones) adequate.
  - Annual budgets include adequate funding for normal collection and compilation of balance of payments statistics.
  - Budgets supplemented with additional resources for specific improvement projects.
  - 2007: project planned on balance of payments and IIP data application processes to improve processes and streamline supporting computer platforms.

### 0.2.2 Measures to ensure efficient use of resources are implemented

- Performance assessment and management
  - Annual staff assessments based on teamwork, individual contribution, and achievement of predefined objectives; revised to align with salary mobility and performance bonuses.
  - Executive performance assessed on achievement of objectives aligned with the three-year strategic plan and its annual update.
  - CBCH has a sophisticated budget formulation and budget execution monitoring process to enhance resource allocation and achievement of results.

- Process improvement initiatives
  - Projects in pipeline to assess and streamline work procedures for improved efficiency.

### 0.3       Relevance

- 0.3.1 Monitoring relevance and practical utility
  - Users can make searches and suggestions via the bank’s website (http://www.bcentral.cl/esp/tools/sugerencia.htm).
  - Information available on periodicity and timeliness (http://www.bcentral.cl/esp/infoeconomica/publicacion).
  - DBPDE conducted user interviews in November 2005 and 2006 to obtain feedback on quality, timeliness, and coverage; no formal user groups exist.
  - Recommendation: Set up a CBCH statistics users group to provide a permanent forum for discussing ways of better meeting user needs.
  - High sensitivity and capacity to respond to user needs from CBCH units, including research and management.
  - Examples of responsiveness:
    - Dissemination of supplementary information on CBCH balance sheet and international reserves (September 2006).
    - Preparation of a quarterly IIP series for research use, to be publicly disseminated with a methodological document in the first half of this year.
    - 2007 project developments in response to user requests: holders of resident debt securities, financial account and income statement series, direct investment series with geographical breakdown.

- External engagement
  - DBPDE participates in regional (CEMLA and others) and extra-regional (OECD) meetings and seminars; active in IMF Committee on Balance of Payments Statistics.
  - CBCH Statistics Committee meets once a month to assess requirements and new needs.

### 0.4 Other quality management

- 0.4.1 Processes to focus on quality
  - High-impact plans for 2006-2008; DBPDE involved in specific 2007 projects (project leader or participant):
    - Documentation of methods and processes used to produce statistics.
    - Impact of Valuation Change Reports (IVV) on the measurement of exports.
    - Study on services exports by mode of supply.
    - Foreign direct investment series with a geographical breakdown.
    - Financial account and income statement series for the 1985-1995 period with categories consistent with the Balance of Payments Manual, fifth edition, (BPM5).
    - Holders of resident debt securities (jointly with the DCIE).
    - Application of processes for balance of payments and IIP (jointly with the DCIE).
  - Achievement of these objectives is binding on DBPDE management.
  - Training programs to enhance staff skills and improve statistical quality.

- DCIE strategic initiatives and budget for the year
  - Application of processes for development of balance of payments and IIP statistics (jointly with DBPDE).
  - Achievement of compliance with international standards for data collection and validation via:
    - Replacement of forms for measuring trade credits.
    - Incorporation of portfolio investment forms into the CIER.
    - Holders of resident debt securities (jointly with DBPDE).
    - Measurement of family remittances (jointly with DBPDE).
  - Annual performance objectives include activities aimed at strengthening efficiency and improving data quality considering international standards and reporter cost.

- 0.4.2 Processes to monitor quality of the statistical program
  - Initiatives to improve quality through specific projects using the Data Quality Assessment Framework (DQAF); IMF mission ROSC recommendations expected to inform prioritization.
  - Analyses of history of different balance of payments versions, and level/sign of errors or omissions:
    - 2004: database of reviews of main balance of payments accounts to assess frequency statistics (number of reviews and lags).
    - Detailed analysis using volatility and coverage indicators for exports subject to IVVs; findings informed review timetables.
    - 2006: statistical study on reviews (see 4.3.3) concluding intermediate versions could create distortions and gross component indicators performed well in time and relative to the Euro zone.
    - 2006: project to evaluate errors or omissions (see 3.4.3).
  - Documentation of processes started in 2006: fact sheets detailing sources and methods at most disaggregated level and assessing quality of each data piece to assign reliability levels and inform improvements.
  - DCIE and DBPDE have established permanent data validation commitments subject to continuous revision.

- 0.4.3 Processes to incorporate quality in planning
  - Projects defined based on comparative assessments of current methods and known best practices, quality improvements from ongoing monitoring, periodic reviews, and user feedback on new data requirements.

### 1. Assurances of integrity

- 1.1 Professionalism

  - 1.1.1 Statistics produced on an impartial basis
    - Professional independence of the BCCH established in Article 1 of its Constitutional Organic Law.
    - No specific statutes preventing external influence on content/diffusion of balance of payments statistics, but in practice there is no external or internal interference affecting results; users perceive this as part of CBCH culture.
    - October 2005 restructuring: GIIE now reports to general management rather than Research Division to strengthen internal independence; internal communication procedures adjusted.
    - End-2006: GIIE and DCIE restructured to focus primarily on compilation, secondarily on collection and validation.
    - Staff up to department management recruited through vacancy announcements and internal/external competitions; no special treatment for balance of payments staff.
    - Promotion policy: open competition within Bank to promote mobility; promotions also via reclassification validated by audit of posts.
    - Training provided through courses, seminars, internships; DBPDE staff turnover low, supporting specialization.

  - 1.1.2 Choices of sources, techniques, and dissemination informed solely by statistical considerations
    - Balance of payments compiled by DBPDE according to professional criteria and with complete independence.
    - CIER: Organic Law stipulates no discrimination in information requests; thresholds established for reporting. Example: for commercial credits, information requested from all exporters and importers with annual f.o.b value equal to or more than $5 million.

    - Data dissemination follows clearly defined procedures and commitments in timing, media and content.

  - 1.1.3 Right to comment on erroneous interpretation and misuse
    - DBPDE provides necessary clarifications in both cases but does not comment publicly.
    - Comments are reserved for the GIIE—which may consult the DBPDE if necessary—in response to a prior request from the Board.

*Source: _cr07318 - 0.2       Resources*

### 1.2       Transparency

### 1.2       Transparency

### Terms and conditions for statistics collection, processing, and dissemination
- The terms and conditions under which balance of payments statistics are prepared are stipulated in Article 53 of Organic Constitutional Law 18840.
- The law may be obtained in print version from the Publications Department of the CBCH (copies are also available in English).
- Information is also available to the public on the terms of the agreement establishing specific obligations regarding the coverage and timeliness of the information published (http://www.bcentral.cl/esp/normativa/acuerdos/pdf/1311-03-061221.pdf).

### Internal governmental access prior to release
- The balance of payments is released simultaneously to all users outside the CBCH, with the exception of the Ministry of Finance, which has access to the data 24 hours before publication.
- The public is informed of this by a footnote in the SDDS metadata.

### Identification of statistical products
- The balance of payments data released to the public are not identified as a product of the DBPDE but rather of the institution.
- Technical notes circulated through the publication Chilean Economy identify the authors of articles; examples including contributions from DBPDE staff are Chilean Economy, volume 7, Nº 3: Measurement of Foreign Direct Investment in the Balance of Payments, and Economic Research Series Nº 46: Effects of Valuation of the IIP of Chile.

### Advance notice of major methodological changes
- Users are informed of any major change in methodology, source data, or statistical techniques simultaneously with the publication of the data in the new form depending on the magnitude of the changes.
- Notification methods include notes or other texts, press releases, press conferences, and/or briefings with analysts.
- Historical examples: inclusion of reinvested earnings on direct investment in 1998, and the adoption of the BPM5 and the dissemination of the IIP in mid-2002.

---

### 1.3       Ethical standards

### Staff behavior guidelines
- There is a Manual of Ethical Principles (March, 2003) that establishes clear standards describing procedures when potential conflict of interest situations arise (Section II, 1).
- The manual includes clear ethical standards on measures to be adopted to prevent the misuse or misinterpretation of statistics (Section II, 2).
- The manual includes standards on:
  - Incompatibilities and conflicts of interest.
  - Use and confidentiality of information.
  - Communication (with the press and public appearances).
  - Participation in seminars and conferences.
  - Use of the work of the CBCH.
  - Acceptance of gifts and invitations.
  - Service fees.
  - Use of the name and image of the CBCH.
- A special regulation exists for the handling of highly sensitive information (HSI).
- The manual is available on the CBCH’s Intranet and new staff are made aware of it in orientation courses organized by the Human Resources Directorate.

---

### 2.0       Methodological soundness

### 2.1 Concepts and definitions
- The current, capital and financial accounts of the balance of payments are broadly in conformity with the BPM5 guidelines.
- Double-entry system applied to the extent possible; net residual embedded in the errors and omissions item.
- Distinctions:
  - Clear distinction between the income component and the goods and services components.
  - In the financial account, transactions in assets and transactions in liabilities are recorded separately.
- Resident institutional units are defined in conformity with the BPM5 (center of economic interest in Chilean economic territory). Examples:
  - Enterprises, including unincorporated branches, are considered to be residents.
  - For construction enterprises operating as unincorporated establishments, the one-year guideline is applied flexibly with other considerations (e.g., maintain a complete set of accounts, pay income taxes to the host country).
  - Residence of units operating mobile equipment in international waters or airspace is attributed to the operating enterprise’s country of residence.
  - Residence of offshore entities and enterprises operating in free zones is attributed to the economy in which they are located.
  - Residence of Special Purpose Entities (SPEs) is attributed to the economy in which they are located.
  - International organizations and supranational authorities with headquarters in Chile are not considered residents of any national economy.
  - All units of general government, including embassies and military bases, are considered to be residents of Chile.
- Transactions defined according to BPM5, including exchanges and unrequited transfers.
- Specific concept applications:
  - Foreign direct investment transactions are defined by applying the 10 percent of ownership rule, strictly in the case of the outward investment, and in principle, in the case of direct investment in Chile.
  - Reserve assets defined considering monetary authorities’ effective control and availability for use; include gross assets related to the Reciprocal Credit System of the Latin American Integration Association (ALADI) while liabilities are shown as short-term drains in the reserve and international liquidity template. In this system assets and liabilities cannot be separated in the compensation sessions and therefore registries should be made on a net basis (assets net of liabilities).
    - Recommendation: Record under reserve assets the net balance of the Agreement on Reciprocal Payments and Credits of ALADI, that is, A accounts (exports or asset balances) minus B accounts (imports or liabilities balances) in consolidated form for all the member countries of said Agreement.
  - Distinction between inbound and outbound transportation is made at the point after goods are loaded on board at the customs frontier of the exporting country (consistent with f.o.b. valuation).
- Some transactions are not included in the financial accounts where a strict resident/nonresident distinction is not possible; recorded with a broader criterion (transactions in assets issued by nonresidents and held by residents, and liabilities whose debtors are resident in Chile and whose holders are nonresidents).
- Concepts and definitions are regularly reviewed by the DBPDE.

### 2.2 Scope
- Coverage aims to follow BPM5: all resident-nonresident transactions covered (e.g., repairs on goods, nonmonetary gold, purchase of computer software, mineral exploration, e-commerce transactions, leases and other transferable contracts, reinvested earnings, intercompany lending, private sector portfolio investment, trade credit, short-term debt transactions, debt arrears, and noncash transactions).
- In practice, some resident-nonresident transactions are not measured or only partially covered due to lack of information (e.g., goods for processing, shuttle trade, or smuggling—estimated flows are insignificant and not included in national accounts).
- All resident units engaged in transactions with nonresidents are covered, including:
  - Incorporated or unincorporated affiliates of nonresident enterprises.
  - Resident territorial enclaves in the rest of the world (e.g., embassies).
  - Free zones, bonded warehouses and factories operated by offshore enterprises under customs control and within Chilean economic territory. For exports from free zones to the rest of the world, adjustments are made based on estimates of trade margins along with national accounts statistics.
- Securities coverage issues:
  - Securities issued by resident economic units on the domestic market and held by nonresidents may be excluded when not captured by DBPDE surveys. A new project is being formulated to broaden survey coverage.
  - Securities issued abroad by residents and acquired by other residents: information obtained from institutional investors and other residents through DBPDE surveys, which means they can be excluded from external assets and liabilities of the balance of payments and the IIP.

### 2.3 Classification/sectorization
- Institutional units classified according to BPM5. Examples and practices:
  - Government guaranteed external debt transactions are attributed to the institutional sector of the borrower.
  - Public enterprise claims and liabilities are attributed to Other Sectors; Ministry of Public Works bond issues were assigned to the general government.
  - Transactions in goods and services by government entities abroad are classified under government services not-included-elsewhere.
  - Long-term construction projects (more than one year) are classified under the foreign direct investment component.
  - Borrowing and lending between direct investors and direct investment enterprises are classified under Foreign Direct Investment (FDI), except when transactions are between affiliated financial intermediaries. Trade credits between affiliates are excluded from this treatment and recorded under Other Investment, Liabilities for Trade Credits, Other Sectors.
    - Recommendation: Reclassify the trade credit flows granted by the matrices to their affiliates in the case of enterprises included in the directory of FDI enterprises as flows of Direct Investment Liabilities, Other Capital in the case of imports with customs information registered under the “collect mode.”
  - Directional approach inconsistency: only assets are recorded under investment abroad, while under direct investment in Chile, only liabilities are recorded; reverse investments treated inconsistently.
    - Recommendation: Apply the directional principle to foreign direct investment in cases where enterprises resident in Chile invest in their parent companies (reverse investments) using the CBCH’s available information in accordance with BPM5 guidelines.
  - In the case of direct investment in Chile, the 10 percent rule is not strictly applied, but distortionary effect estimated to be small; almost all investments recorded pursuant to Decree Law 600 fall within the more than 10 percent ownership criterion.
  - Loan transactions of the banking sector are classified separately from currency and deposits transactions of this sector.
  - Short-term and long-term attribution of transactions in other investment component made according to original maturity of the financial instrument.
  - Coverage of the capital account is partial: debt forgiveness, grants (capital transfers, based on customs information, as of 2003), and sporadic information on the sale of licenses have been recorded. A survey will be conducted to collect data on sale of licenses.
  - Goods for processing are not measured in the goods item of the balance of payments; feasibility is being looked into of creating new customs statistics codes to gather this information. Transaction type not seen as relevant to Chilean economy.
    - Recommendation: Work with the NCS to agree on a distinction in customs concepts to identify operations involving goods for processing and for repair; CBCH should clarify conceptual criterion to avoid confusion with similar operations (e.g., temporary admission regime is not necessarily linked to goods for processing operations).

### 2.4 Basis for recording

#### 2.4.1 Valuation (market prices)
- Principle of market valuation from BPM5 is used:
  - Total imports and exports valued on an f.o.b. basis.
  - Monetary gold valued at market prices.
  - Financial instruments generally traded on a regular basis valued using market price quotations.
  - Nontradable financial instruments valued at nominal value.
  - Financial derivatives valued at market prices.
- Transactions are generally calculated directly rather than estimated from stock data.
- Short-term borrowing (liabilities) and, for the moment, trade credits are not included. For trade credits, CIER tests are being conducted on a new data source providing consistent information on transactions and balances.
- For other assets and liabilities of the financial account, information is available to estimate valuation and other changes to reconcile initial and final values of the IIP.
- Exports of goods without known market prices at change of ownership use provisional price estimates, subsequently replaced by actual prices. No adjustments are made for transactions between affiliated enterprises.
  - Recommendation: Work with the NCS, and possibly the ITS, to agree on an amendment to the rule that authorizes a lag of up to nine months for exporters to submit the final values of their exports recorded under the “not firm” mode so as to substantially reduce these lead times.
- Transactions are measured in U.S. dollars and generally reported in the same currency:
  - For goods where U.S. dollar is not original currency, conversions use the customs rate for the respective month (valid for one calendar month and fixed by the National Customs Service).
  - For financial account transactions in currencies other than the U.S. dollar, conversions use market rates prevailing on the date of the transaction.
  - International reserves flows are calculated daily using the respective rates.

#### 2.4.2 Recording basis (accrual)
- Change of ownership as specified in BPM5 is the principal recognition and timing rule:
  - Services recorded when rendered (delivered or received).
  - Dividends recorded as of dates payable or dates declared payable.
  - Reinvested earnings on direct investment recorded in the periods when earned.
  - Loan drawings entered at time of actual disbursement; loan repayments entered when due for payment.
  - Interest on the main asset holdings recorded on an accrual basis (international reserves and assets of institutional investors).
- Exceptions to accrual:
  - Goods: main source is customs with no adjustments at time of recording; for imports lodged in free zones prior to final clearance, the first date of entry is used.
  - External loan liabilities and portfolio investment (public and private) interest sometimes recorded at date of effective payment. A 2006 project aimed to incorporate accrual for such interest; results forthcoming to assess impact and decide on definitive incorporation. The DMFAS system version 5.2 includes a module to calculate interest accrual.
    - Recommendation: record transactions related to government external debt liabilities on an accrual basis.
- Pension fund investments in mutual funds abroad follow draft (March 2007 version) of new balance of payments manual (version 6): traditional methodology records only dividends distributed by the mutual fund as income; new treatment imputes income in the current account with a counterpart in the financial account as a “reinvestment” of gains, and reflects in the IIP in the “transactions” column.

#### 2.4.3 Grossing/netting procedures
- Current and capital account transactions are recorded on a gross basis in line with BPM5 recommendations.
- Some financial account transactions are recorded on a net basis (trade credits, assets and liabilities of the banking system, reinvested earnings, currency and deposits and short-term loans), consistent with BPM5 standards; others recorded on a gross basis (direct investment in equity capital and other capital, portfolio investment, financial derivatives, reserve assets and other long-term investment).
- Individual asset and liability components are recorded separately to provide more information.

*Source: _cr07318 - 1.2       Transparency*

### 3. Accuracy and reliability

### 3. Accuracy and reliability

### 3.1 Source data — composition, collection, and core sources
- Core data sources for Chile’s balance of payments:
  - (1) the CIER, (2) customs documents, (3) surveys with different periodicities (mainly for services and direct and portfolio investment in Chile), and (4) estimates.
- The current collection system is mixed (by itemized transaction and consolidated) and has been evolving gradually towards direct reporting by the entity that provides consolidated information.
- CIER developments and coverage:
  - Reformulated in normative terms in April 2001, transitioning from a capital controls system to one aimed mainly at collecting statistics; developed in 2002 with a balance of payments approach.
  - Maintains a system of daily reporting of income and outlays of foreign currency conducted through the banks and exchange bureaus (the formal exchange market).
  - Expanded traditional concept of foreign currency purchases and sales to cover transfers of foreign exchange not associated with purchases or sales and adopted a new coding system.
  - Daily reporting supplemented by multiple forms for direct respondents to approximate balance of payments, external debt and IIP concepts more closely.
- CIER reporting periodicities and respondent types (selected):
  - Resident enterprises engaged in international cargo and passenger transport and representatives of nonresident enterprises (quarterly).
  - Insurance and reinsurance companies (quarterly).
  - Credit and debit card operators (quarterly).
  - Residents making direct, portfolio, or other investments abroad — monthly, quarterly, or annual reports depending on respondent.
  - Residents trading financial derivatives abroad — monthly, and daily for banks only (with respect to foreign exchange).
  - Residents receiving capital or credit from abroad — each time contracted.
  - Merchandise importers and exporters for trade credits (quarterly with a monthly breakdown).
- Additional administrative and external sources used for compilation:
  - Financial statements or other enterprise information (Uniform Coded Datasheets (FECUs), etc.).
  - Regulatory institutions: SPF (pensions), SBFI (banks and financial institutions), SSI (insurance).
  - Monetary and financial statistics on external positions of resident units to validate financial transactions (Monetary and Financial Statistics Manual (IMF. 2000), Chapters V and VII).
  - CBCH accounts for reserve assets to distinguish transactions from valuation changes.
  - Government sources: Ministry of Finance (MOF), national Treasury, Internal Taxation Service (ITS).
  - Administrative records of foreign agencies (international police, Foreign Investment Committee).
  - International sources (BIS statistics) on deposits of nonbanking sector residents in Chile.
  - Estimates.

### 3.1 Source data — surveys used and sampling practices
- Services surveys (periodicities and notes):
  - Travel surveys by SERNATUR — periodicity: monthly, with quarterly reports for 2007.
  - Communications survey for national accounts — periodicity: annual by census.
  - Courier national accounts surveys — periodicity: annual–census.
  - Survey of other land transport to estimate passenger credits — periodicity: every three years.
  - National accounts survey of sea and air transport agencies — periodicity: annual by census.
  - Surveys of royalties and license fees — anchor survey, last conducted in 2000 by sampling.
  - Survey of cultural and recreational services for sporting activities — anchor survey, last conducted in 2000 by sampling.
  - Survey of cultural and recreational services for audiovisual services — anchor survey, last conducted in 2000 by sampling.
  - Survey of government services to measure credits — periodicity: annual by sampling (applied to international agencies).
  - Survey of various business, professional, technical services — periodicity: every two years by sampling.
  - Survey of information technology services to measure debits — periodicity: every two years by sampling.
  - Survey of information services to measure credits and debits — periodicity: every two years by sampling.
  - Surveys of commissions on foreign trade activities to measure credits — periodicity: every three years by sampling.
  - Survey of represented imports to measure credits — periodicity: every three years by sampling.
- Financial account, income and position surveys:
  - Survey of investment — periodicity: annual.
  - Survey of custodian bank portfolios to measure portfolio investment liabilities — periodicity: annual.
- Planned/other projects (as stated):
  - Interinstitutional technical group preparing an exploratory survey on services exports including the four modes of supply defined in GATS, to be applied in 2007.
  - Special project in 2007 to define a methodology for measuring remittances, with an exploratory survey to measure credits and debits.
- Direct investment in Chile survey:
  - Sample selection: forced inclusion of companies with largest accumulated investments and inclusion of companies generating significant earnings.
  - Questionnaires include automatic validations of form and consistency; can be answered online; data downloaded into a standardized database.
  - Annual evaluation to update directory (changes in name, organizational structure, new investments, changes in investment ownership) and assessment based on response rate (in percent) compared with previous years.
  - Instructions for respondents include multiple methods (including mass meetings); surveys include a section for respondents' views and suggestions.
  - Use of validation criteria and auxiliary data (FECU) to minimize nonsampling errors.
  - Recommendation: Reach an agreement with the SSI for enterprises to report using the FECU, including a breakdown by residents and nonresidents.

### 3.1 Source data — directories, coverage, reporting thresholds, and validation
- Services survey directories and sample design:
  - Sample divided into two strata: forced inclusion stratum (enterprises with highest transaction volumes) and a sample stratum (small enterprises).
  - Selections for services survey samples evaluated every one, two, or three years.
  - Directories compiled from ITRS background data (including tax code at individual level for transactions above a threshold), list of services exporters reported to Customs, Industrial Property Department directory, Economic Activity Classifier directory, and ITS information.
  - No services directories available for the NSI to date; work underway within the exploratory services exports survey.
  - Procedures exist for frequently integrating new transactor units into registers and accounting for mergers, cessations, and other changes (activities of the new DCIE).
- Reporting instructions and thresholds:
  - Reporting instructions request information on all purchases/sales and transfers, requiring that any transaction over USD10,000 or associated with a transaction over USD10,000 be properly coded for balance of payments purposes.
  - Instructions for reporting annual data request information on transactions above a given reporting threshold for verification and adequate representation.
  - The largest unclassified transactions executed through the banks are fully checked on a monthly basis to ensure they do not correspond to balance of payments transactions.
  - Transactions below the threshold are estimated using information available on exchange operations received through the banks.
  - No sample survey of banks and nonbanking institutions with accounts abroad is conducted to classify low-value transactions across balance of payments components; unclassified transactions channeled through banks are exhaustively checked monthly.
- ITRS and bank reporting:
  - Coverage of reporting banks in the ITRS is integrated with bank records prepared especially for the CBCH.
  - Structure of reporting for banks with accounts abroad is consistent with banks’ accounting practices, allowing extraction of individualized data from operational records.
  - Main classification problem arises when economic transactors must indicate source or intended use of foreign exchange and when administrative bank personnel must interpret classification.
  - ITRS report forms are processed electronically, pilot-tested, and can be provided over the Internet.
  - The system serves as a directory for assessing compliance, implementing surveys, and establishing transaction classification.
- Validation and cross-checks:
  - Forms received through CIER are checked against financial statements for largest companies that submit balance sheets to the superintendency.
  - Different data sources used to validate survey methodology (e.g., coefficients derived from foreign exchange codes, customs data, ITS for professional and IT services).
  - Data from administrative records used to derive balance of payments components are adequate, and their gaps in detail and coverage are known and accounted for.
  - Meetings are held with financial market participants and the business community to identify new developments affecting compilation (example: meetings with banks on new accounting standards effective beginning of 2007; meetings with exporters/importers when new trade credit reporting forms introduced in early 2007).
  - The financial press is monitored to identify international transactions and transactors for coverage expansion when significant capital movements occur.
  - International standards, guidelines, and practices are monitored for changes affecting compilation.

### 3.1 Approximations to BPM5 concepts, adjustments, and timeliness
- Consistency with BPM5 and methodological requirements:
  - Source data drawn from surveys and the ITRS are consistent with definitions, scope, classifications, valuation, and time of recording employed by the DBPDE.
  - Pilot tests conducted to ensure data collected allow compilation according to international guidelines.
  - Administrative records and other external sources provide reasonable approximations of methodological requirements.
  - Adjustments are made to customs statistics coverage to approximate balance of payments concepts (e.g., inclusion of items not traditionally counted in customs exports; incorporation of restricted imports and other items).
  - Private external debt balances information comes mostly from CIER and published DMFAS data, both consistent with BPM5 recommendations.
- Timeliness:
  - Data collection and processing timetables adequately meet recommended standards of timeliness and periodicity for disseminating balance of payments statistics.
  - Some preliminary figures are published based on estimates from indirect indicators due to source periodicity or deadlines (examples):
    - Transport enterprise forms have such a lag they can only be incorporated in the balance of payments one quarter later.
    - Annual survey forms are generally entered in the second round of production of the year-end balance of payments.
  - Respondents are made aware of reporting deadlines.
  - A different CBCH unit (DCIE, through DCIE and DBD departments) is responsible for collecting data from core sources and employs rigorous follow-up to ensure timely receipt.
  - Goods account source data from customs are practically 100 percent available online and received by CBCH collection units at the time sent by Customs.
  - Updated directories of institutional investors and banking institutions are maintained by superintendencies to guarantee optimum data coverage.
  - Online reporting is available with corresponding notification and validation systems.
- Legal framework note and recommendation:
  - An improvement in the legal framework to ensure that the private sector is required to provide information for statistical purposes would help bolster efforts to provide coverage of this sector’s transactions.
  - Footnote reference: See recommendation in 0.1.4.

*Source: _cr07318 - 3. Accuracy and reliability*

### 3.2 Assessment of source data

### 3.2 Assessment of source data

### Source data monitoring and quality control
- Source data (censuses, sample surveys, administrative records) are routinely assessed for coverage, sample error, response error, and nonsampling error; results of the assessments are monitored and made available to guide statistical processes.
- Information collected on nonsampling errors (survey operations, biases, misclassification, processing, and nonresponse) is monitored and analyzed regularly to reduce these errors.
- Questionnaire changes are assessed for their effects on estimation; questionnaires are intended to be user friendly and should not vary greatly between applications.
- Source data are analyzed in the context of revisions; major changes based on updated data during the revision cycle are incorporated to avoid accumulation of revisions over extended periods.
- Outliers and atypical differences in periodic responses are identified by users; extreme values are confirmed with respondents and records maintained on confirmations. Respondents to services surveys are contacted directly to confirm their responses.
- In surveys for the financial accounts, when validation information exists and data do not tally, the best quality information is kept and atypical values are confirmed with respondents.
- Source data are analyzed for underreporting/misreporting, including checks for temporal consistency and consistency with related data sources.
- Information about nonsampling errors is monitored, including undercoverage, misclassification, processing, and nonresponse.
- Total purchases and sales data are cross-checked with banking institutions' accounts.
- Discrepancies in ITRS transactions that are not of balance of payments nature are carefully analyzed (e.g., type of transaction, spread, and sign of fluctuation); the “Other” code is reviewed extensively.
- High-value transactions are confirmed with respondents; transactions exceeding the US$100,000 threshold are reviewed in detail and, if there are doubts, respondents are consulted.
- Accuracy of administrative records and other sources is checked regularly against additional secondary sources (example: FECU data compared with survey and report form data; SERNATUR definitions, ITS classifications, and Treasury figures checked against public information on general government).

### Statistical techniques and data compilation (3.3)
- Compilation procedures minimize processing errors such as coding, editing, and tabulation errors.
- Goods data are validated by the receiving unit, which checks spreadsheets for inconsistencies in codes and units of measure; squares (cuadraturas) of values are revised and errors in prices are detected via a price checker (depurador de precios) comparing reported prices with historical prices under the same tariff heading; detected errors are corrected online as soon as caught.
- Adjustments to unit records are made only when clearly warranted; unusual values are not replaced or modified unless clearly required and can be identified in datasets.
- Procedures for imputation and adjustment for nonresponse are soundly based; ancillary and benchmark information are used where appropriate (example: measurement of direct investment uses surveys or forms, and where no response exists other available data such as exchange rates are used).
- Example adjustment: direct investment in Chile’s mining sector measures growth in profits by applying the price of copper to information on production volumes.
- When reported values are out of range or improbable, adjustments replace them with information from other sources or with estimates.
- Estimation techniques adjust data for missing observations by analyzing trends in recent information, links with related indicators; used regularly to supplement transport and insurance information or to make initial estimates when there is a lag in receipt of source data.
- Adjustments for undercoverage and/or population units out of scope are based on other sources or related indicators.
- Uncoded ITRS data are individually researched if significant and reassigned to specific balance of payments components; remaining uncoded data are not attributed and are concluded to largely correspond to transactions between residents.
- Discrepancies between sums of quarterly data and annual survey data are removed through benchmark procedures or indicators.

### Other statistical procedures and specific adjustments (3.3.2)
- Specific procedures developed to adjust data from various sources to improve coverage, classification, and valuation and to conform to BPM5 guidelines include:
  - The c.i.f./f.o.b. factor in customs declarations for valuing imports reported in c.i.f. values only (free zones, etc.).
  - Adjustments to record payments for services, including taxes withheld at source, using ITS information.
  - Adjustments to SENATUR’s travel estimates to account for definition differences (e.g., residence); CBCH will receive data starting in 2007 that will cover the entire year and reduce need for period estimates.
  - Net transactions in telecommunication services reported in the ITRS are adjusted to a gross basis using information collected from a sample of telecommunications firms and quarterly estimates based on indirect indicators.
  - Prices of major exports initially reported by Customs at provisional values are adjusted to approximate market values.

### Assessment and validation of intermediate data and statistical outputs (3.4)
- Intermediate results are validated against other information sources where applicable:
  - ITS data on withholding taxes on income are used to assess accuracy of reported income transactions.
  - Financial press information is used to verify high-value direct investment and other transactions (example: a very large and complex direct investment transaction in 2006 was checked against press information); a system tracks transactions weekly, including multi–fiscal-year projects, from origin through final compilation of balance of payments, external debt, and IIP.
- Statistical discrepancies in intermediate data are assessed and investigated by routine comparisons:
  - Investment income payments and receipts are regularly assessed against corresponding stock data in the IIP.
  - Freight earnings are assessed relative to the value/volume of trade flows.
  - Reported financial flow data are reconciled with changes in corresponding stock data for external debt and other IIP elements; since February 2005, published IIP information includes crosschecks of initial and final stocks and flows, valuation changes, identifying price and parity effects and other changes.
  - Discrepancies between merchandise trade and associated financial flows from the ITRS are reviewed.
- Statistical discrepancies and other potential indicators of problems are investigated:
  - Staff monitor errors and omissions and seek to understand them via cross-checks between current, capital, and financial accounts.
  - A 2006 project examined trends in the residual account to find causes of negative values recorded in recent years; outcomes included improved coordination with data compilers, training, compliance with delivery deadlines, and careful monitoring via press and ITRS.
  - Despite increasing flows, net errors have remained at acceptable levels: over the last 11 years the ratio of errors and omissions to the sum of current account credits and debits has remained at about 2 percent.
  - Selected external debt transactions are compared with Joint BIS-IMF-OECD-World Bank Statistics on external debt or BIS locational international banking statistics.
  - Chile contributed to the IMF comparative study of short-term external debt statistics (Guide for Compilers and Users of External Debt Statistics, Part II, Chapter 14, Country Experiences: Chile, 14.6 4 – 14.79).
  - Balance of payments and IIP results are checked against the Coordinated Portfolio Investment Survey.
  - Bank information is compared with BIS data; BIS is the basic data source for external deposits of nonbanking institutions.

### Revision studies (3.5)
- Studies and analyses of revisions are carried out routinely and used internally to inform statistical processes.
- A largely descriptive revision study in 2004-2005 quantified number of versions and magnitude of revisions.
- In 2006 a revisions project calculated internationally developed data quality indicators measuring reliability, relevance, timeliness, and consistency.
- The 2004-2005 study informed a revisions policy announced to the public in May of 2006 together with figures for end of first quarter of that year.
- Results of revision studies were used to improve price estimates for exports.
- Since 2006, causes of revisions are described publicly and a database on revisions was created, providing access to all published versions of each balance of payments account and the magnitudes involved.

### Revision measurement methods and indicators (4.3.3)
- Revisions are quantified and explained in CBCH’s Monthly Bulletin and on the website as part of notes accompanying end-of-quarter and end-of-year balance of payments releases.
- Internal studies assess reliability of preliminary data and the revision cycle; the most recent (2006) introduced measurement by statistically reliable international indicators.
- MAPE (Mean Absolute Percentage Error) measures reliability as the average size in percent of the revision (sum of the differences between versions) using gross variables.
- MARE (Mean Absolute Relative Error) considers variability of average standardized revisions (standard deviation) using net variables.
- Comparative results: for the current account Chile has relatively better MARE levels compared with the euro area; for income and investment accounts the European figures are higher; with respect to MAPE, results applied to goods and services are much the same for the euro area and Chile.
- Directional reliability (measuring whether final-version changes show same trend as preliminary estimates) generally produces very good results, generally over 70 percent.

### Periodicity and timeliness (4.1)
- Periodicity of dissemination for some indicators exceeds SDDS requirements:
  - Goods transactions: weekly.
  - Financial account flows and the international reserves and liquidity template: monthly.
  - Reserve assets flows: bi-weekly.
- Timeliness: Quarterly balance of payments statistics are disseminated within three months after the end of the reference period, as established in the SDDS.
  - Balance of payments for the first three quarters of each year is published within the first two months after the end of the reference period (May 23, August 23, and November 23).
  - Preliminary data for the end of the fourth quarter are published within three months (March 23 of each year).
  - This timeliness is consistent with the IIP and external debt figures.

### Consistency (4.2)
- Concepts, definitions, and classifications for quarterly and annual balance of payments statistics are the same as those used to compile annual estimates; summary quarterly publications for the current year are less detailed than annual presentation.
- Sum of quarterly statistics is consistent with annual statistics.
- Financial transactions data are reconcilable with changes in the IIP; a reconciliation table (transactions, exchange rate, price, other changes) is disseminated every six months.
- Consistent balance of payments time series are available from 1996.
- When source data, methodology, or technique changes are introduced, historical series are reconstructed as far back as reasonably possible (implementation of BPM5 in 2002 recalculated series back to 1996; project under development to extend series to1990).
- Methodological notes explain main breaks and discontinuities; time series are presented in blocks of comparable periods and notes are included in other publications (for example, CBCH’s Monthly Bulletin).
- Comments describing trends and explaining unusual changes are included in quarterly and annual publications (preliminary data) in CBCH’s Monthly Bulletin and on CBCH’s website.
- Balance of payments statistics are consistent with national accounts; balance of payments is a source for national accounts compilation, and submission deadlines and cross-checking requirements have been established to promote consistency.
- Banking sector transactions in balance of payments are consistent with monetary and financial statistics; procedures exist to check monetary figures reported for balance of payments purposes and reserve asset balances can be reconciled distinguishing transactions, prices, and exchange rate effects.
- Balance of payments components that include external debt data are consistent with respective debt balances, except for bond liabilities:
  - For external debt the bond liabilities are recorded at nominal value.
  - In the balance of payments and the IIP they are recorded at market value.
  - In both cases figures are adjusted for repos.
- Recommendation: Facilitate reconciliation of the IIP and external debt statistics, in particular portfolio investment liabilities issued by the government, given that different criteria are used (market and nominal values, respectively).

### Revision policy and transparency (4.3)
- Revisions follow an established, reasonably stable annual procedure.
- In May 2006 a note in CBCH’s Monthly Bulletin explained the revision policy and timetable for balance of payments figures for that year; the same was done in March 2007 for 2007 figures (note published in March Monthly Bulletin and on CBCH website).
- Notes on revisions generally describe reasons for the revision cycle; an internal document details main reasons for the revision schedule.
- Comments accompanying end-of-quarter and end-of-year data include a section quantifying and explaining significant changes.
- Revision policy allows for revisions outside the regular cycle where necessary; such revisions are explained in public notes.
- Preliminary and revised data are clearly identified: first quarterly data are indicated as preliminary when published; users are informed at dissemination whether published data are revised.

*Source: 3.2 Assessment of source data (from the provided IMF content unit).*

### 5.         Accessibility

### 5.         Accessibility

### 5.1 Data accessibility

- Statistics presentation and layout
  - Balance of payments statistics are disseminated according to the standard components of the BPM5, and with time series.
  - A summary version of the balance of payments is distributed quarterly and the more detailed version annually, based on the standard BPM5 components.
  - Additional series are disseminated with various levels of detail (disaggregation) to meet users' needs.
  - The summary quarterly versions of Chile’s balance of payments for the current year are accompanied, in some cases, by more detailed information for analytical purposes.
  - The statistics are published clearly with tables to facilitate analysis.
  - The statistical database on the CBCH’s website is able to plot graphs of the different series, and to transfer the series to electronic spreadsheets.
  - Comments on events in the current reporting period are included in the note accompanying the end-of-quarter data, which is published in the monthly bulletin and on the CBCH’s website.
  - Seasonally adjusted series for goods imports are published monthly.

- Dissemination media and format
  - End-of-quarter and end-of-year data on the CBCH’s website are highlighted by a short report entitled Informe de Coyuntura (http://www/bcentral.cl/esp/inforeconomica/coyuntura/).
  - Press releases are drafted with highlighted comments (example: press release of March 23, 2007 in the Press Room and Current Affairs sections: http://bcentral.cl/esp/prensa/informacion/pdf/23032007a.pdf).
  - More complete and detailed balance of payments and external debt statistics than those published quarterly on the web and in the CBCH’s Monthly Bulletin are disseminated in the Annual Reports (Balance of Payments of Chile and External Debt of Chile), published in hard copy, on electronic media (balance of payments) and in PDF format on the CBCH’s website.
  - Recent data on the longest time series can be accessed free of charge through an electronic database maintained by the CBCH entitled Base de Datos Estadísticos, and in electronic spreadsheet format, Estadísticas Económicas at (http://www/bcentral.cl/esp/infoeconomica/seriesindicadores/index_se.htm).

- Release schedule, simultaneity, and access
  - A schedule announces in advance the dates on which the statistics are to be released; that schedule is published in the Monthly Bulletin on the CBCH’s website.
  - The statistics are released punctually, that is according to the preannounced schedule.
  - The public is informed of the statistics being released and of the procedures to access them (e.g., Internet, publications).
  - The statistics are made available to all interested users simultaneously.

- Non-routine dissemination upon request
  - In addition to routinely disseminated statistics, other general statistics are made available when requests are made in compliance with the requirements established in a procedure agreed by the CBCH Board in October 2005: http://www.bcentral.cl/esp/normativa/acuerdos/2005.htm.

### 5.2 Metadata accessibility

- Documentation and methodological transparency
  - The methodological bases and data sources are presented in the Balance of Payments Yearbook, available in printed and electronic form and in PDF format on the CBCH’s website.
  - Important methodological changes are explained when the information is published, either by means of explanatory notes, or directly (example: when the revision of a series that included reinvested profits was made public, a press conference was called and a meeting with expert analysts was held).
  - As a result of a project on documentation being carried out in 2007, a document will be prepared for external users that will update and broaden the documentation on the sources and methods used for compiling Chile’s balance of payments.
  - The SDDS metadata, methodological summaries relative to the SDDS, and other related descriptions are reviewed and updated annually and can be consulted on the CBCH’s website (http://www.bcentral.cl/esp/infoeconomica/fminedd/ecofinantialdata/).

- Recommendation (verbatim)
  - It is recommended that the version of the methodological notes on the balance of payments and the IIP and the Chile’s Annual Balance of Payments Report be included in a separate, easily identifiable manner in the Statistical Publications section of the CBCH’s website.

- Levels of detail for audiences
  - More specialized information documents have been prepared and made public (research paper entitled Measurement of Foreign Direct Investment in the Balance of Payments, in the series Economía Chilena, Vol. 7, No. 3, December 2004).
  - A methodological document on the calculation of external trade deflators, a study on balance of payments revisions, and one on a quarterly IIP series are in the final stages of preparation.

### 5.3 Assistance to users

- Contact points and user support
  - Prompt and knowledgeable service and support are available to users of statistics. The CBCH’s webmaster forwards any questions not sent to a specific addressee.
  - The SDDS website, from which the CBCH’s webpage can be accessed, identifies contact points for each topic for enquiries by mail, telephone, facsimile, or by e-mail.

- Catalogs and charging information
  - Publication catalogs exist in hardcopy and on the CBCH’s website (http://www.bcentral.cl/esp/publ/index.htm).
  - The prices of these products are clearly indicated to the public.

### Key statistics and table references preserved from the source

- Table 11. Chile: Data Quality Assessment Framework (July 2003): Summary of Results for Balance of Payments Statistics (Compiling Agency: Central Bank of Chile) is presented in the source, including:
  - Assessment elements and symbols with Key to symbols: NA = Not Applicable; O = Practice Observed; LO = Practice Largely Observed; LNO = Practice Largely Not Observed; NO = Practice Not Observed; SDDS = Complies with SDDS Criteria.
  - Specific findings and comments captured in the table include, among others:
    - 0.1 Legal and institutional environment: comment — Improve the legal framework to allow the CBCH to request information from the private sector.
    - 0.3 Relevance: comment — There is no users’ group.
    - 2.3 Classification/sectorization: comment — Reclassify trade credit flows of enterprises included in the directory of FDI enterprises as FDI liability flows in the case of inputs with custom information registered under the “collect mode.” Follow the BPM5 in applying the directional criteria to the reverse investment transactions of resident FDI enterprises using the CBCH’s available information.
    - 2.4 Basis for recording: comment — Record transactions related to interest on public and private external debt on an accrual basis.
    - Various elements marked O, LO, or X as presented in the table.

- Appendix I: Summary of the Special Data Dissemination Standard (SDDS) (as presented)
  - Data dimension: dissemination of 18 data categories, including component detail, covering the four main sectors (real, fiscal, financial, and external) of the economy, with prescribed periodicity and timeliness.
  - Access dimension:
    - dissemination of advance release calendars providing at least one-quarter advance notice of approximate release dates, and at least a one-week advance notice of the precise release dates; and
    - the simultaneous release of data to all users.
  - Integrity dimension: 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 the occasion of statistical release; provision of information about revision and advance notice of major changes in methodology.
  - Quality dimension: dissemination of documentation on statistical methodology and sources; dissemination of component detail and/or additional data series that make possible cross-checks and checks of reasonableness.
  - SDDS subscribers are required to post metadata on the IMF’s Dissemination Standards Bulletin Board (DSBB) and maintain an Internet NSDP linked to the DSBB.
  - IMF monitoring is limited to coverage, periodicity, timeliness, and dissemination of advance release calendars. Source: http://dsbb.imf.org

*Content based on "5. Accessibility" from the supplied IMF document.*

### 4.1 Periodicity and

### _cr07318 - 4.1 Periodicity and

### Serviceability: periodicity, timeliness, consistency, and revisions
- Periodicity and timeliness follow internationally accepted dissemination standards.
- 4.1.2 Timeliness follows dissemination standards.
- 4.2 Consistency—Statistics are consistent within the dataset, over time, and with major datasets.
  - 4.2.1 Statistics are consistent within the dataset.
  - 4.2.2 Statistics are consistent or reconcilable over a reasonable period of time.
  - 4.2.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks.
- 4.3 Revision policy and practice—Data revisions follow a regular and publicized procedure.
  - 4.3.1 Revisions follow a regular and transparent schedule.
  - 4.3.2 Preliminary and/or revised data are clearly identified.
  - 4.3.3 Studies and analyses of revisions are made public (see also 3.5.1).
- Serviceability summary statement: "Statistics, with adequate periodicity and timeliness, are consistent and follow a predictable revisions policy."

### Accessibility: data, metadata, and assistance
- 5.1 Data accessibility—Statistics are presented in a clear and understandable manner, forms of dissemination are adequate, and statistics are made available on an impartial basis.
  - 5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts).
  - 5.1.2 Dissemination media and format are adequate.
  - 5.1.3 Statistics are released on a preannounced schedule.
  - 5.1.4 Statistics are made available to all users at the same time.
  - 5.1.5 Statistics not routinely disseminated are made available upon request.
- 5.2 Metadata accessibility—Up-to-date and pertinent metadata are made available.
  - 5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from internationally accepted standards, guidelines, or good practices are annotated.
  - 5.2.2 Levels of detail are adapted to the needs of the intended audience.
- 5.3 Assistance to users—Prompt and knowledgeable support service is available.
  - 5.3.1 Contact points for each subject field are publicized.
  - 5.3.2 Catalogs of publications, documents, and other services, including information on any charges, are widely available.

### Users’ survey: design and participation
- Survey administered with assistance from the Central Bank of Chile (CBCH) and the National Statistics Institute (NSI).
- Questionnaire developed by IMF staff, conducted by the NSI, targeting users of macroeconomic statistics.
- Questionnaires sent: 207.
- Responses received: 50.
- Respondent types included: academia, financial institutions, international organizations, the media, government agencies, embassies, producers’ associations, and consulting companies.
- Six meetings with key users of Chile’s macroeconomic data organized by the CBCH complemented the survey.

### Key findings from users and mission assessment
- General satisfaction:
  - Most users were satisfied with methodological soundness, timely dissemination, accessibility, and frequency of official macroeconomic statistics.
  - About 50 percent of respondents felt that methodological concepts and analytical presentations were clearly described.
- Areas for improvement identified by users:
  - PPI lacks credibility and is not used much.
  - Coverage and detail of GFS and PPI seen as insufficient.
  - Some users unsure whether an early release calendar for GFS existed and whether it was met.
  - Information on revisions for NA, PPI, GFS, and BOP could be improved.
  - Access to metadata on NA, PPI, and GFS was found somewhat difficult.
  - Demand for more detail and longer time series on NA, BOP, PPI, IMACEC (monthly index of economic activity), and GFS.
  - Strong demand for data on household income distribution, average prices by industry, volume of sales by industry, and reliable and consistent series on employment statistics.
  - Need to improve coverage and consistency of labor statistics emphasized across meetings.
- Agencies cited:
  - CBCH cited as the agency with the most reliable information.
  - NSI website cited as very useful because detailed data are available on line.
- Users requested more information on revision policies and expanded breakdowns of national accounts, price, government finance, and balance of payments statistics.

### Questionnaire results (selected aggregates and exact counts)
- Table 12: Questionnaire Results Analyzed by Type of User (March, 2007)
  - Total Sent: 207.
  - Total Received: 50.
  - In % Received: 24.2.
  - By type (Total Sent / Total Received / In % Received):
    - Government: 20 / 6 / 30.0
    - Financial institutions: 45 / 20 / 44.4
    - Embassies and international organizations: 7 / 2 / 28.6
    - Universities, researchers, and consultants: 94 / 14 / 14.9
    - Media: 29 / 1 / 3.5
    - Supervisory, regulatory institutions: 7 / 2 / 28.6
    - Private nonfinancial enterprises: 1 / 1 / 100.0
    - Producers’ associations: 4 / 4 / 100.0
    - Total: 207 / 50 / 24.2
- Table 13: Results of Chile’s User Survey (March 2007) — Selected question totals
  - 15. Which official statistics do you use regularly? (Total Number of Responses)
    - a. National accounts (NA): 38
    - b. Prices: 48
    - c. Government finance statistics (GFS): 35
    - d. Monetary statistics (Monetary): 32
    - e. Balance of payments (BOP): 30
    - f. Other: 38
      - Production indices: 31
      - Labor market: 38
      - Merchandise trade: 20
      - International reserves and foreign currency liquidity: 21
      - External debt: 24
      - International investment position: 15
      - Other: 3
  - 16. Where do you obtain the official statistics?
    - a. Official press releases and publications on macroeconomic statistics: 46
    - b. Private sector summaries and analyses: 20
    - c. Official policy papers: 31
    - d. Publications from international organizations about the country: 23
    - e. Other sources: 5
  - 17–18. Refer to official descriptions of sources and methods used to compile official statistics?
    - Yes: 36
    - No: 12
  - 19. Purposes of use (counts)
    - a. Analysis of current developments for short-term decision making: 32
    - b. Analysis of trends for longer-term policy formulation: 31
    - c. Econometric model building and forecasting: 34
    - d. Economic research: 31
    - e. Comparison with economic developments in other countries: 20
    - f. General economic background: 19
    - g. Other: 4
  - 20. Coverage and detail — "In general, are you satisfied with the coverage of official statistics?" (Yes / No by domain totals)
    - NA: Yes 34 / No 8
    - Prices: Yes 40 / No 9
    - GFS: Yes 23 / No 14
    - Monetary: Yes 33 / No 3
    - BOP: Yes 33 / No 4
    - Other: Yes 5 / No 3
  - 20. Coverage and detail — "In general, are you satisfied with ... level of detail?" (Yes / No)
    - NA: Yes 30 / No 12
    - Prices: Yes 38 / No 10
    - GFS: Yes 21 / No 17
    - Monetary: Yes 31 / No 4
    - BOP: Yes 27 / No 8
    - Other: Yes 3 / No 6
  - 21. Periodicity and timeliness — Frequency satisfaction (Yes / No)
    - NA: Yes 36 / No 3
    - Prices: Yes 44 / No 1
    - GFS: Yes 29 / No 8
    - Monetary: Yes 36 / No 1
    - BOP: Yes 31 / No 4
    - Other: Yes 4 / No 1
  - 21. Periodicity and timeliness — Timeliness satisfaction (Yes / No)
    - NA: Yes 33 / No 6
    - Prices: Yes 45 / No 2
    - GFS: Yes 27 / No 11
    - Monetary: Yes 35 / No 3
    - BOP: Yes 32 / No 4
    - Other: Yes 7 / No 1
  - 22. Other dissemination practices — Knowledge of a publicly disseminated release calendar (Yes / No)
    - NA: Yes 26 / No 18
    - Prices: Yes 36 / No 13
    - GFS: Yes 15 / No 27
    - Monetary: Yes 25 / No 13
    - BOP: Yes 26 / No 12
    - Other: Yes 3 / No 3
  - 22. If a calendar exists, are statistics released on announced dates? (Yes / No)
    - NA: Yes 26 / No 0
    - Prices: Yes 37 / No 1
    - GFS: Yes 15 / No 5
    - Monetary: Yes 28 / No 0
    - BOP: Yes 26 / No 0
    - Other: Yes 3 / No 0
  - 22. Is there enough information about revisions to official statistics? (Yes / No)
    - NA: Yes 23 / No 15
    - Prices: Yes 29 / No 15
    - GFS: Yes 18 / No 20
    - Monetary: Yes 25 / No 9
    - BOP: Yes 23 / No 11
    - Other: Yes 1 / No 1
  - 23. Accessibility — Can you easily access the official statistics? (Yes / No)
    - NA: Yes 36 / No 6
    - Prices: Yes 44 / No 5
    - GFS: Yes 29 / No 13
    - Monetary: Yes 37 / No 2
    - BOP: Yes 35 / No 2
    - Other: Yes 7 / No 2
  - 23. Accessibility — Can you easily access information pertaining to official statistics you use (explanatory notes, methodological descriptions)? (Yes / No)
    - NA: Yes 22 / No 15
    - Prices: Yes 34 / No 11
    - GFS: Yes 18 / No 20
    - Monetary: Yes 25 / No 11
    - BOP: Yes 25 / No 10
    - Other: Yes 5 / No 3
  - 23. Is the methodology information sufficiently clear and adequate? (Total Yes / No)
    - Yes 18 / No 18
  - 23. How do you get access to official statistics? (counts)
    - Official releases: 22
    - Hard copy publications: 30
    - Data specifically requested: 10
    - Official website: 49
    - Other: 2
    - E-mail requests: 9
  - 24. Overall assessment — Underlying methodology sound and appropriate? (Yes / No)
    - NA: Yes 34 / No 4
    - Prices: Yes 36 / No 7
    - GFS: Yes 31 / No 8
    - Monetary: Yes 34 / No 2
    - BOP: Yes 32 / No 2
    - Other: Yes 4 / No 2
  - 24. Overall assessment — Official statistics unbiased and accurate? (Yes / No)
    - NA: Yes 34 / No 5
    - Prices: Yes 38 / No 6
    - GFS: Yes 35 / No 4
    - Monetary: Yes 36 / No 2
    - BOP: Yes 33 / No 3
    - Other: Yes 3 / No 2
  - 24. Comparison with other countries in the region (Better / Same / Worse)
    - NA: Better 25 / Same 11 / Worse 0
    - Prices: Better 30 / Same 9 / Worse 3
    - GFS: Better 25 / Same 7 / Worse 4
    - Monetary: Better 23 / Same 10 / Worse 0
    - BOP: Better 25 / Same 8 / Worse 0
    - Other: Better 2 / Same 0 / Worse 0
  - 24. Overall quality rating (1 poor to 5 excellent) — Average rating:
    - NA: 3.9
    - Prices: 4.0
    - GFS: 3.7
    - Monetary: 4.3
    - BOP: 4.2
    - Other: 3.6
  - Note: NA = National Accounts; Prices refers to: CPI (Consumer Price Index) and PPI (Producer Price Index); GFS = Government Finance Statistics; Monetary = Monetary Statistics; and BOP = Balance of Payments Statistics.

### Users’ detailed comments (summarized verbatim points)
- Demand for more disaggregated and household-level data:
  - "Information of income and its distribution, and information on households and dwellings are published at aggregated level." (Respondent 1)
  - "The last household budget survey is more than 10 years old." (Respondent 13)
  - Requests for hours worked, duration of unemployment, term of contracts, payment of indemnifications per years of service, household wealth and its composition, breakdown of household consumption expenditure. (Respondent 16)
- Labor and employment statistics:
  - "A common denominator in all the meetings was the need to improve the coverage and consistency of labor statistics."
  - Calls for reliable and consistent series on employment statistics and publication of underemployment (voluntary and involuntary). (Respondent 12, 17)
- National accounts, IMACEC, inventories, and investment by sector:
  - Need for statistics on inventories and investment by sectors; more breakdowns on the IMACEC; improvement in regional series and links between old and new series. (Respondents 7, 8, 12, 17)
- Prices and CPI concerns:
  - CPI base year outdated and limited coverage (Great Santiago). (Respondent 5)
  - Doubts about CPI methodology, imputation issues (e.g., transport cost of Transantiago), and unclear CPI revisions. (Respondents 6, 12, 17)
  - Request for prices of monthly exports and imports of other basic products (salmon, steel, wood, wine, etc.). (Respondent 17)
- Government finance statistics (GFS) and transparency:
  - Need for more explanations and methodology on GFS; GFS only covers central government; need for a functional classification of government expenditure. (Respondents 2, 9)
  - Concerns about transparency and access to budgetary topics, public enterprises, and use of large resources (e.g., from copper). (Respondent 11)
  - Difficulty knowing whether an early release calendar for GFS existed and was met.
- Balance of payments (BOP) details and volatility:
  - Need for more detailed balance of payments statistics by sector, enterprise, and region. (Respondent 4)
  - Financial services figures in BOP vulnerable to assumptions on reinvested earnings and dividends, presenting noticeable revisions affecting Direct Foreign Investment figures. (Respondent 7)
- Data dissemination and methodological documentation:
  - NSI has detailed online data but regional detail often delayed; CBCH data considered reliable.
  - Users want more information on revision policies and clearer methodological descriptions. (Multiple respondents)
  - Calls for seasonally adjusted data, official backward links of indices and long series, and greater budgetary allocation for the NSI. (Respondent 17)
- Aggregation and detail compared internationally:
  - Statistics considered too aggregated compared with practices in the United States (e.g., price indices by type of good and service; longer series; greater level of detail in tax data). (Respondent 15)
- Miscellaneous:
  - Need for data on volume of sales and average prices by industry. (Respondent 3)
  - Citizens' participation and statistics showing details on minorities requested. (Respondent 18)
  - Some users indicated institutional dissemination irregularities (e.g., Ministry of Finance and Treasury publications). (Respondent 10)

*Source: _cr07318 - 4.1 Periodicity and (excerpt), IMF staff report materials and users’ survey results as provided in the source content.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2007/_cr07318.pdf_