## _cr04398

## Source details

**Canonical URL:** [_cr04398](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04398.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04398.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04398.pdf.json)

---

### Executive overview
- Review of Pakistan’s data dissemination practices against the IMF’s General Data Dissemination System (GDDS) and in-depth assessment of quality of national accounts, consumer and wholesale price indices, and government finance, monetary, and balance of payments statistics.
- Data producers: Federal Bureau of Statistics (FBS), Ministry of Finance (MOF), State Bank of Pakistan (SBP).
- IMF Statistics Department mission visited Islamabad and Karachi during December 1-16, 2003.
- Authorities committed to international standards (GDDS participation); shortcomings identified that could detract from accurate and timely analysis and policy formulation.
- Recommendations prioritized national accounts and government finance statistics as greatest urgency.
- Some recommendations addressable with current resources; SDDS subscription within a year would require additional resources and forceful implementation of an action plan.

### GDDS participation and SDDS subscription: coverage, periodicity, timeliness
- Pakistan meets GDDS recommendations for coverage, periodicity, and timeliness of most categories, notable exceptions:
  - Timeliness of GDP:
    - First GDP releases published about two weeks before the end of the fiscal year and based on nine to ten months’ observed data and two to three months’ projections.
    - First annual GDP data based on 12 month’s observations disseminated with a lag of 11 months (vis-à-vis recommended 6-9 months).
    - Publication of quarterly national accounts (in preparation) would remedy this.
  - Lack of annual data on wages/earnings.
- SDDS subscription requires dissemination of:
  - biweekly data on the analytical accounts of the SBP (note: beginning in February 2004, biweekly data on the analytical accounts of the SBP are disseminated on the website and the Statistical Bulletin);
  - monthly transaction data on central government operations;
  - quarterly data on the GDP, employment, unemployment, and wages/earnings;
  - more detailed breakdown of data on central government debt;
  - annual international investment position;
  - update and expand metadata on compilation and dissemination practices.
- By developing and forcefully implementing an action plan, Pakistan should be able to meet remaining SDDS requirements within a year.

### Prerequisites of quality (legal, institutional, resources, quality awareness)
- Legal/institutional:
  - General Statistics Act No. 69 of 1975 establishes framework; does not explicitly state FBS responsibility for national accounts or prices but implied by tradition.
  - National Statistical Council (NSC) exists but has not been active in practice; National Accounts Committee (NAC) coordinates national accounts work.
  - Confidentiality guaranteed; fines for noncompliance not indexed and largely insignificant.
- Resources:
  - National accounts: NAW staffing 61 positions with 10 vacancies; computing resources deficient (many obsolete 386/486 PCs); base-year estimates at 1980/81 prices.
  - Price statistics: central staff 22 (12 statistical officers, 6 statistical assistants); regional/field offices often lack PCs; price program budget broadly sufficient though travel allowances insufficient.
  - GFS: MOF staff and computing adequate for limited GFS but AGPR constrained by lack of computer automation.
  - SBP: financial and computing resources adequate; 14 of 76 Statistics Department staff primarily responsible for monetary statistics; plans to equip staff with Pentium IV computers noted.
- Quality awareness:
  - FBS convenes NAC and standing panels; conducts revision studies and methodological reviews.
  - MOF quality awareness weaker; GFS compilers lack user perspective.
  - SBP demonstrates quality awareness via “Vision Statement,” Business Plan, and Planning and Development Division.

### Integrity and transparency
- Professionalism and ethical guidelines present across FBS, MOF, SBP.
- FBS:
  - NAC reviews and approves accounts; FBS monitors press and provides clarifications.
  - Terms/conditions for data collection summarized on website and GDDS metadata; internal government pre-release access is partially publicized.
- MOF/GFS:
  - Longstanding convention of MOF independence in compiling and publishing fiscal data; limited public information on production/dissemination arrangements.
- SBP:
  - Practices encourage impartial compilation; no internal government access prior to publication; SBP requests attribution and publicizes methodological changes.

### Methodological soundness — dataset highlights
- National accounts:
  - Concepts follow System of National Accounts 1968 (1968 SNA); scope very narrow and does not meet minimum requirements for 1993 SNA.
  - FBS instituting comprehensive revision and expects to disseminate revised estimates and quarterly accounts during first half of 2004.
  - Recording: market output valued at factor cost; sales taxes excluded from intermediate consumption; government transactions recorded on cash basis.
  - Important exclusions: own-account production, informal activities, many intangible assets.
- CPI/WPI:
  - CPI scope limited to 52 urban centers (covering about 68 percent of urban population); WPI does not cover small-scale manufacturing enterprises (FBS response disputes and asserts WPI covers small-scale manufacturing).
  - CPI and WPI concepts broadly follow international guidelines; COICOP not in use for CPI.
  - Housing in CPI estimated using smoothed construction-cost–based indices (geometric mean smoothing over the past 24 months); housing weight in CPI 23.4 percent.
- GFS:
  - Broadly based on GFSM 1986; treatment of privatization proceeds resembles GFSM 2001.
  - Scope covers budgetary general government but excludes extra-budgetary units; classification departs substantially from GFSM 1986; recordings largely cash basis.
- Monetary statistics:
  - Analytical framework reflects IMF’s draft Guide to Money and Banking Statistics (December 1984); SBP plans to adopt MFSM.
  - Sectorization and instrument classification in published tables not fully MFSM-compliant; repos treated as outright sales.
- Balance of payments:
  - Compiled broadly in accordance with BPM5 concepts but presentation on web/publications still BPM4; SBP plans to adopt BPM5 presentation.
  - Residency exceptions: EPZ enterprises and offshore banks regarded as nonresidents; Pakistani nationals abroad with dwellings in Pakistan considered residents.

### Accuracy and reliability — source data and techniques
- National accounts:
  - Source data rely on benchmark censuses, HIES 2000/01, administrative sources; CMI every five years; business register ~56,000 corporations but likely incomplete.
  - Techniques weak: estimates compiled at 1980/81 prices; fixed ratios from old benchmarks; no comprehensive procedures to estimate non-observed economy; inventory valuation adjustments and perpetual inventory method not fully applied.
  - Revision studies: FBS commissioned 23 studies (first May 2001, last May 2002) used in rebasing to 1999/2000.
- CPI/WPI:
  - Family Budget Survey 2000-01 used for weights; Family Budget Survey results not published though FBS indicates next survey will be published.
  - Price collection practices detailed; sampling and verification procedures in place (supervisors verify 10 percent of recorded price quotations for CPI; 15-20 percent for WPI).
  - Housing measurement method departs from international best practice; suggestion to replace construction-cost–based rental estimates.
- GFS:
  - Source data from AGPR/AGs timely and from SBP for financing; compilation directly from accounting data.
  - Discrepancy between deficit/surplus and financing exists; corrections posted in period discovered not original period.
  - No systematic reconciliation with other sources; no systematic revision studies.
- Monetary statistics:
  - Main source weekly aggregated statements lacking required breakdowns; monthly report forms provide greater detail and are recommended for compilation.
  - Manual entry of weekly data creates error potential; SBP initiated data warehouse and transition to month-end compilation (beginning June 2004 noted).
  - Validation procedures adequate with cross-checks; revision studies not routine.
- Balance of payments:
  - Main sources: ITRS, enterprise reports, administrative records; generally timely and aligned with international definitions.
  - Statistical techniques generally sound; revision studies not carried out.

### Serviceability — relevance, timeliness, periodicity, consistency, revisions
- General:
  - Disseminated data broadly meet users’ needs in detail but not in coverage; informal user survey during mission cited.
  - Inter-sectoral consistency (GFS, BOP, MS) needs strengthening.
- Timeliness and periodicity examples (preserve numeric values exactly):
  - National accounts: first annual GDP based on 12 month’s observations disseminated with a lag of 11 months; first provisional releases published about two weeks before end of fiscal year based on nine to ten months’ observed data and two to three months’ projections; estimates compiled annually (GDDS) but quarterly accounts in preparation.
  - CPI: Pakistan timeliness 11 days (GDDS timeliness 1-2 mths); GDDS periodicity M; Pakistan periodicity M.
  - Broad money and credit aggregates: Pakistan periodicity M; Pakistan timeliness 1 mth (GDDS timeliness 1-3 mths).
  - Balance of payments aggregates: Pakistan periodicity Q; Pakistan timeliness 2 mths (GDDS timeliness 6 mths).
  - Central govt. debt: Pakistan periodicity Domestic debt: M / External debt: Q; Pakistan timeliness Domestic debt: 2 mths / External debt: 1 qtr.
  - Gross official reserves denominated in U.S. dollars: Pakistan periodicity W; Pakistan timeliness 1 wk.
- Revision policies:
  - FBS: provisional estimates updated six months after end of period (submitted only to government agencies); comprehensive revision released eleven months after end of period; revision cycle not publicized.
  - CPI/WPI: revisions follow schedule; baseyear changed every ten years (reduction to five years under consideration); CPI release on the 11 th of month following reference month.
  - SBP: no formal revision policy historically; SBP plans to inform users of revision policy and include documentation in publications.

### Accessibility — data, metadata, user assistance, release practices
- Opportunities to expand accessibility across datasets; official statistics should be made available to all users simultaneously.
- Advance release calendars not always provided; hardcopies released to selected government users and press shortly before website publication.
- Metadata:
  - GDDS metadata posted on DSBB; SBP provides hyperlink to GDDS metadata.
  - Comprehensive sources-and-methods document for monetary statistics exists internally (Working Manual); public metadata limited.
- User assistance and contacts:
  - Assistance needs improvement across agencies; FBS and SBP provide contact information on websites but not always in hardcopy publications; MOF contact limited on websites.
  - FBS publications and website list names and email addresses but not subject fields of contacts.
- Dissemination sites cited:
  - Federal Bureau of Statistics: http://www.statpak.gov.pk
  - State Bank of Pakistan: http://www.sbp.org.pk
  - Ministry of Finance: http://www.finance.gov.pk
- Note: beginning June 2004 SBP improved dissemination of analytical accounts based on month-end data with a lag of two weeks and published an advance release calendar.

### Data dissemination arrangements and main producers
- FBS: national accounts, prices, other economic statistics.
- SBP: monetary and balance of payments statistics.
- MOF: statistics on central and general government operations and central government debt.

### Plans for improvement (selected)
- Real sector: expand data sources, broaden scope, implement 1993 SNA recommendations, change base year to 1999/2000 and rebasing cycle to five years.
- Prices: introduce COICOP; develop PPI with base year 1999/2000; review housing treatment in CPI.
- Labor market: FBS investigating feasibility of quarterly employment/unemployment and dissemination of wages/salaries; plan to increase LFS sample from about 21,000 households to 32,000 (8,000 households per quarter) to produce representative quarterly estimates (planned use of regular resources in 2005-06).
- Government finance statistics:
  - Fiscal Responsibility Law submitted to parliament for approval; Project for the Improvement of Financial Reporting and Auditing (PIFRA).
  - MOF plans to develop and disseminate metadata and an advance release calendar for GFS; intends to adopt GFSM 2001 in the longer term.
- Monetary statistics:
  - SBP to adopt MFSM guidelines, disaggregate nonfinancial corporations and households within private sector, collect data on nonbank financial institutions, and develop metadata.

### Mission conclusions (key points summarized)
- GDDS participation in place; SDDS subscription requires specific dissemination and metadata improvements.
- Urgent attention needed on national accounts and government finance statistics.
- Principal prerequisites: institutional strengthening (autonomy for FBS), clearer responsibility for comprehensive GFS, more resources and training, stronger coordination.
- Enhance transparency of dissemination practices and ensure simultaneous public access to strengthen integrity and accessibility.

### DQAF application and summary assessment (selected numeric and coded findings preserved)
- DQAF dimensions assessed for six datasets: national accounts (NAS), consumer price index (CPI), wholesale price index (WPI), government finance statistics (GFS), monetary statistics (MS), balance of payments statistics (BOPS).
- Selected DQAF summary entries (preserve codes):
  - 0.1 Legal and institutional environment: NAS = LO; CPI = O; WPI = O; GFS = LO; MS = O; BOPS = O.
  - 0.2 Resources: NAS = LO; CPI = LO; WPI = LO; GFS = LO; MS = LO; BOPS = O.
  - 0.3 Quality awareness: NAS = O; CPI = O; WPI = O; GFS = LO; MS = O; BOPS = O.
  - 1.1 Professionalism: NAS = LO; CPI = O; WPI = O; GFS = O; MS = O; BOPS = O.
  - 2.1 Concepts and definitions: NAS = LO; CPI = O; WPI = O; GFS = LO; MS = LO; BOPS = LO.
  - 3.1 Source data: NAS = LNO; CPI = O; WPI = LO; GFS = O; MS = LNO; BOPS = O.
  - 3.5 Revision studies: NAS = O; CPI = O; WPI = O; GFS = LO; MS = O; BOPS = LNO.
  - 4.2 Timeliness and periodicity: NAS = LO; CPI = O; WPI = O; GFS = O; MS = LO; BOPS = O.
  - 5.1 Data accessibility: NAS = LO; CPI = LO; WPI = LO; GFS = LO; MS = LO; BOPS = O.
- Plans recorded in DQAF:
  - NA: disseminate quarterly GDP data by activity and category of expenditure in early 2004; change base year to 1999/2000 (short term).
  - GFS: MOF plans to disseminate monthly GFS with a lag of one month (short term) and migrate to GFSM 2001 (medium term).
  - MS/BOP: SBP plans to adopt MFSM and BPM5 presentation formats and compile month-end data (short term); additional metadata to be published.

### Staff recommendations — General (IV. STAFF RECOMMENDATIONS)
- Priority: high; national accounts and government finance statistics urgent.
- General recommendations:
  - Pursue initiative to give more autonomy to FBS and conduct public awareness exercises.
  - Improve survey response rates (e.g., amend General Statistics Act to increase fines).
  - Review statistical policies/practices with users, addressing presentation, metadata, advance release calendars, simultaneous release, publicity of unpublished data, conducting revision studies, and providing detailed contact information.
  - Provide more training in statistical methodologies and modernize computer equipment.
  - Strengthen coordination among institutions to enhance intersectoral consistency.
  - Update and enhance GDDS metadata on DSBB regularly.

### Staff recommendations — Dataset‑specific (selected, verbatim)
- National accounts:
  - Amend NAC role so final dissemination decision rests with FBS.
  - Introduce regular (annual) establishment surveys; broaden economic activities covered.
  - Improve timeliness: disseminate estimates revised for government agencies six months after end of reference period.
  - Implement procedures to estimate non-observed economy.
- CPI/WPI:
  - Implement planned introduction of COICOP as soon as possible.
  - Review housing treatment in CPI; replace construction-cost rental estimates with purchase/maintenance cost index.
  - Consider expanding CPI scope to all urban areas and eventually rural areas.
- Government finance statistics:
  - Set up a small group within MOF dedicated to compilation and dissemination of GFS.
  - Migrate to GFSM 2001 classification structure as far as possible.
  - Provide classificatory detail specified in GFSM 2001 and publish time series of discrete quarterly data.
  - Examine statistical discrepancy between deficit/surplus and financing each quarter under Fiscal Monitoring Committee.
- Monetary statistics:
  - Compile and disseminate monetary statistics based on monthly report forms rather than weekly statements.
  - Enhance data validation to ensure mutual claims between SBP and scheduled banks recorded consistently.
  - Implement MFSM guidelines on sectorization, classification, accounting; incorporate accrued interest into outstanding instrument amounts.
  - Note: Beginning in June 2004 monetary statistics based on month-end data and more in line with MFSM.
- Balance of payments statistics:
  - Compile external transactions of EPZ and offshore bank enterprises and treat them as residents.
  - Disaggregate counterpart of exports of crude oil under product sharing agreements into (i) income component (direct investment income) and (ii) repatriation of direct investment capital (financial account FDI).
  - Strengthen estimation of reinvested earnings to include all earnings capitalized in reserve accounts.
  - Reclassify government-guaranteed loans to sector of beneficiary of loan proceeds (recommendation adopted in 2004).
  - Complete compilation of international investment position on BPM5 basis.

### Authorities' response — selected actions (Response by the Authorities, November 29, 2004)
- FBS:
  - Proposal to merge statistical bodies into a single autonomous organization submitted to Government; expected completion during 2005.
  - Advance release calendar on FBS website; data simultaneously released and posted; contact information (names, e-mail, phone, fax, postal address) provided on publications and website.
  - Project proposals for capacity building with World Bank and training with UNFPA negotiated.
  - Plans to replace outdated 386/486 PCs; seek external support for regional/field offices and National Accounts.
  - Plans for supply and use tables and institutional sector accounts on annual basis; rebasing to 1999/2000 and subsequent five-year rebasing cycle agreed (next rebasing year 2005-06).
  - Quarterly national accounts and quarterly employment estimates planned; LFS to increase sample to 32,000 households to yield quarterly estimates starting 2005-06.
- SBP:
  - Began disseminating monetary statistics more in line with MFSM in June 2004; public advance release calendar available.

### CPI and WPI — key numeric and procedural facts preserved
- CPI:
  - Scope: 52 urban centers covering 68 percent of urban population.
  - Housing weight in CPI: 23.4 percent.
  - Price and weight reference period: full year 2000-2001.
  - CPI release schedule: released on the 11 th of the month following the reference month.
  - CPI base-year practice: changed every ten years historically; reduction to every five years being considered.
  - Sample/HIES: Family Budget Survey conducted in 2000-01; total selected households 41,060; sample per quarter ~10,000 households; LFS annual sample ~21,000 households (plan to increase to 32,000).
- WPI:
  - Basket: 106 commodities and 425 items.
  - WPI coverage excludes about 20 percent of domestically traded goods by WPI definition.
  - WPI release schedule: released on the 11th of the month following the reference month.
  - Supervisors verify each month 15-20 percent of recorded WPI price quotations.

### National accounts — key numeric items preserved
- Base year for constant price estimates prior to rebasing: 1980/81.
- FBS NAW staffing: 61 staff positions with 10 vacancies.
- Business register: about 56,000 records.
- Census of Manufacturing Industries response rate (2000/01): 46 percent of 9,800 factories surveyed.
- Supply of HIES 2000/01: total selected households 14,912 from 1,050 PSUs; 62 percent urban households.
- FBS commissioned 23 revision studies (first May 2001, last May 2002) used to rebase to 1999/2000.

### GFS — key numeric/timeliness facts preserved
- Quarterly central and general government aggregates disseminated within two months after end of reference period.
- Federal budget data supplied within 15 days of end of reference period.
- Audited final accounts available about seven months after end of reference year.
- GFS time series compiled from 1996/97 onward.
- Financing data from SBP provided monthly but with a lag of two months.

### Monetary statistics — key numeric/timeliness facts preserved
- SBP Statistics Department: 76 staff; 14 with primary responsibility for monetary statistics.
- Weekly balance sheet data of SBP available 10 days after reference week; Statistics Department requires 2-3 days to generate analytical accounts from weekly data.
- Monthly report form reporting date: 20th day after end of reference month; Statistics Department requires about 7-10 days to finalize analytical accounts for scheduled banks.
- Weekly aggregated data for scheduled banks provided by Banking Supervision Department 10 days after reference week (Saturday).
- SBP planned/implemented changes:
  - In February 2004 SBP published monetary survey for December 2003 on website.
  - Beginning in June 2004, SBP disseminates monetary statistics largely in line with MFSM and based on month-end data.
- Timeliness benchmarks:
  - Monthly data for SBP and scheduled banks disseminated within four weeks after reference period (GDDS timeliness).
  - SBP end-June fiscal year data historically available with two to three month lag; SBP decided to use provisional data to improve timeliness.

### Balance of payments — key practices and timeliness
- Main source: ITRS from authorized dealers (more than 1,700 ADs).
- BOP compiled on accrual basis; quarterly BOP disseminated within one quarter after reference period (meets SDDS); annual BOP within three months after end of reference period.
- BOPD plans to present BOP in BPM5 format beginning January 2004 internally and July 2004 publicly.
- Revisions: predefined revision practice; preliminary monthly BOP compiled; revisions up to one year; no routine revision studies but major revisions explained.

### Users’ views (informal user survey during mission)
- Survey: 112 contacted, 52 responded (response rate 46 percent).
  - Respondent composition: 35 percent private sector; 25 percent embassies and international organizations; 17 percent public sector agencies; 19 percent academics; 4 percent media.
- Overall quality rating on five-point scale: 3.16.
- Main user criticisms:
  - Coverage gaps (national accounts, production indices, services sector, provincial breakdowns).
  - Periodicity/timeliness weaknesses (national accounts annual and lagged; need for quarterly GDP).
  - Accessibility/metadata/revision policy deficits (absence of advance release calendar, inadequate methodological notes).
  - Calls for improved coordination among agencies and stronger FBS reforms.

*Source: _cr04398 - Executive Summary.*

### Executive Summary ......................................................................................................

### _cr04398 - Executive Summary ......................................................................................................

### Executive overview
- This ROSC provides a review of Pakistan’s data dissemination practices against the IMF’s General Data Dissemination System (GDDS), complemented by an in-depth assessment of the quality of national accounts, consumer and wholesale price indices, and government finance, monetary, and balance of payments statistics.
- Data are produced by the Federal Bureau of Statistics (FBS), the Ministry of Finance (MOF), and the State Bank of Pakistan (SBP).
- Assessments were conducted by an IMF Statistics Department mission that visited Islamabad and Karachi during December 1-16, 2003.
- Authorities demonstrate strong commitment to international standards (GDDS participation), but shortcomings exist that could detract from accurate and timely analysis and policy formulation.
- Recommendations were made to enhance adherence to international statistical standards; national accounts and government finance statistics recommendations have the greatest urgency.
- Some recommendations can be addressed with current resources; SDDS subscription within a year would require additional resources and forceful implementation of an action plan.

### GDDS participation and SDDS subscription: coverage, periodicity, timeliness
- Pakistan meets GDDS recommendations for coverage, periodicity, and timeliness of most data categories, with notable exceptions:
  - Timeliness of GDP:
    - First GDP releases are published about two weeks before the end of the fiscal year and are based on nine to ten months’ observed data and two to three months’ projections.
    - The first annual GDP data based on 12 month’s observations are disseminated with a lag of 11 months (vis-à-vis the recommended 6-9 months).
    - Publication of quarterly national accounts (in preparation) would remedy this situation.
  - Lack of annual data on wages/earnings.
- For SDDS subscription Pakistan needs to disseminate:
  - biweekly data on the analytical accounts of the SBP (note: beginning in February 2004, biweekly data on the analytical accounts of the SBP are disseminated on the website and the Statistical Bulletin);
  - monthly transaction data on central government operations;
  - quarterly data on the GDP, employment, unemployment, and wages/earnings;
  - more detailed breakdown of data on central government debt;
  - annual international investment position;
  - update and expand the metadata on compilation and dissemination practices.
- By developing and forcefully implementing an action plan, Pakistan should be able to meet the remaining requirements for SDDS subscription within a year.

### Prerequisites of quality
- Legal and institutional framework is broadly adequate and will be strengthened by initiative to provide more autonomy to the FBS.
- Institutions demonstrate awareness of quality.
- Responsibility for compilation and dissemination of comprehensive government finance statistics is not defined, hampering compilation of fiscal aggregates with broader coverage than budgetary government.
- Resources are broadly adequate for current programs; additional resources, targeted training, and more formal coordination among statistical agencies are needed for further development and SDDS subscription.
- Coordination among and within institutions compiling related statistics should be strengthened to promote consistent data across statistical systems.

### Integrity
- Institutions demonstrate professionalism, transparency in practices and policies, and provide ethical guidelines.
- National Accounts Committee (with government and private sector participants) reviews and approves accounts, focuses on updated source data and may recommend revisions before dissemination.
- Longstanding convention of independence in MOF compiling and publishing fiscal data, but terms and conditions under which data are compiled and disseminated are not publicized.
- Procedures on government internal access to data prior to public release are not always disseminated.

### Methodological soundness
- Methodologies for price indices and balance of payments broadly follow international standards.
- Significant shortcomings exist in:
  - scope of national accounts;
  - classification and sectorization systems for government finance statistics;
  - classification and sectorization systems for monetary statistics.

### Accuracy and reliability
- Price indices and fiscal statistics are broadly accurate.
- Important shortcomings in data sources for national accounts and monetary statistics.
- Weaknesses in statistical techniques hinder assessment and validation of national accounts.
- Studies of revisions are carried out for all datasets and used to inform statistical processes, except for balance of payments statistics.

### Serviceability (user needs, periodicity, timeliness, consistency)
- Disseminated data broadly meet users’ needs in level of detail but not in coverage (results of an informal user survey are referenced).
- Ongoing consultation with users, especially private sector, should be more formal and regular for government finance, monetary, and balance of payments statistics.
- Significant room for improving inter-sectoral consistency of government finance, balance of payments, and monetary statistics.
- Periodicity and timeliness are generally satisfactory for GDDS, though:
  - periodicity of GDP data should be improved;
  - timeliness of the analytical accounts of the SBP for the end of the fiscal year should be improved (beginning in June 2004, timeliness of the analytical accounts of the SBP ha been improved).

### Accessibility
- Opportunities exist to expand accessibility of macroeconomic data and metadata across most datasets.
- Official statistics should be made available to all users simultaneously.
- Assistance to users needs improvement in all statistical agencies, notably for government finance statistics by providing sector-specific contact information on websites and in hardcopy publications.
- Advance release calendars are not always provided; hardcopies of data are released to selected users within government and to the press shortly before publication on the FBS website.

### Data dissemination arrangements and publications
- Macroeconomic statistics mainly produced by:
  - Federal Bureau of Statistics (FBS): national accounts, prices, other economic statistics;
  - State Bank of Pakistan (SBP): monetary and balance of payments statistics;
  - Ministry of Finance (MOF): statistics on central and general government operations and central government debt.
- Internet dissemination sites cited for public access:
  - Federal Bureau of Statistics: http://www.statpak.gov.pk
  - State Bank of Pakistan: http://www.sbp.org.pk
  - Ministry of Finance: http://www.finance.gov.pk

### Plans for improvement (selected)
- Real sector statistics: expand data sources and broaden scope.
- Prices: introduce COICOP and complete development of a PPI with base year 1999/2000.
- Labor market: FBS investigating feasibility of compiling quarterly employment/unemployment data and disseminating wages/salaries.
- Government finance statistics:
  - Fiscal Responsibility Law submitted to parliament for approval;
  - Project for the Improvement of Financial Reporting and Auditing (PIFRA);
  - MOF plans to develop and disseminate metadata and an advance release calendar for GFS;
  - In the longer term, MOF intends to adopt GFSM 2001.
- Monetary statistics: SBP will continue to adopt the Monetary and Financial Statistics Manual (MFSM) and plans to:
  - disaggregate data on nonfinancial corporations and households within the private sector;
  - collect data on nonbank financial institutions to expand coverage of monetary aggregates;
  - develop and disseminate metadata on compilation of monetary statistics.

### Mission conclusions (key points summarized)
- GDDS participation is in place; SDDS subscription requires specific additional dissemination and metadata improvements.
- Urgent attention needed on national accounts and government finance statistics.
- Institutional strengthening (autonomy for FBS), clearer responsibility for comprehensive GFS, more resources and training, and stronger coordination are principal prerequisites for improvement.
- Enhancing transparency of dissemination practices and simultaneous public access are needed to strengthen integrity and accessibility.

*Source: _cr04398 - Executive Summary.*

### 17.      Regarding the external sector, the SBP plans to (i) continue the full implementation

### _cr04398 - 17.      Regarding the external sector, the SBP plans to (i) continue the full implementation

### SBP plans for the external sector
- Continue the full implementation of the Balance of Payments Manual, fifth edition (BPM5).
- Initiate plans to compile and publish data on the international investment position.
- Disseminate the metadata for the template on International Reserves and Foreign Currency Liquidity on the SBP website.
- Develop advance release calendars for monetary and balance of payments statistics.

### Overview of current practices: coverage, periodicity, and timeliness (GDDS comparison)
- Real Sector: National Accounts
  - Coverage: Yes (meets GDDS)
  - GDDS periodicity: A
  - Pakistan periodicity: A
  - GDDS timeliness: 10-14 months
  - Pakistan timeliness: Final estimates: 11 months.
- Fiscal Sector: Central govt. operations / General govt. operations / Central govt. debt
  - Coverage: Yes / Yes / Yes
  - GDDS periodicity: A / A / A
  - Pakistan periodicity: Q / Q / Q
  - GDDS timeliness: Domestic debt: M; External debt: Q; 6-9 mths
  - Pakistan timeliness: 2 mths / 2 mths / 2 mths
- Financial Sector: Broad Money Survey
  - Coverage: Yes
  - GDDS periodicity: M
  - Pakistan periodicity: M
  - GDDS timeliness: 2-3 mths
  - Pakistan timeliness: 1 mth
- External Sector: Balance of Payments
  - Coverage: Yes
  - GDDS periodicity: A
  - Pakistan periodicity: Q
  - GDDS timeliness: 6-9 mths
  - Pakistan timeliness: 2 mths

Key selected data-category items and Pakistan practice (numeric values preserved exactly as in source):
- GDP (nominal and real): Coverage Yes; GDDS periodicity A; Pakistan periodicity A; GDDS timeliness 6-9 mths; Pakistan timeliness 6-9 mths
- Gross national income, capital formation, saving: Coverage Yes; GDDS periodicity A; Pakistan periodicity A; GDDS timeliness 11 mths; Pakistan timeliness 11 mths
- Manufacturing or industrial production index/indices: Coverage Yes; GDDS periodicity NA; Pakistan periodicity M; GDDS timeliness As relevant; Pakistan timeliness 6-12 wks
- Consumer price index: Coverage Yes; GDDS periodicity M; Pakistan periodicity M; GDDS timeliness 1-2 mths; Pakistan timeliness 11 days
- Employment: Coverage Yes; GDDS periodicity A; Pakistan periodicity A; GDDS timeliness 6-9 mths; Pakistan timeliness 6 mths
- Unemployment: Coverage Yes; GDDS periodicity A; Pakistan periodicity A; GDDS timeliness 6-9 mths; Pakistan timeliness 6 mths
- Central govt. budgetary aggregates: Coverage Yes; GDDS periodicity Q; Pakistan periodicity Q; GDDS timeliness 1 qtr; Pakistan timeliness 2 mths
- Interest payments: Coverage Yes; GDDS periodicity Q; Pakistan periodicity Q; GDDS timeliness 1 qtr; Pakistan timeliness 2 mths
- Central govt. debt: Coverage Yes; GDDS periodicity A; Pakistan periodicity Domestic debt: M / External debt: Q; GDDS timeliness 1-2 qtrs; Pakistan timeliness Domestic debt: 2 mths / External debt: 1 qtr
- Broad money and credit aggregates: Coverage Yes; GDDS periodicity M; Pakistan periodicity M; GDDS timeliness 1-3 mths; Pakistan timeliness 1 mth
- Central bank aggregates: Coverage Yes; GDDS periodicity M; Pakistan periodicity M; GDDS timeliness 1-2 mths; Pakistan timeliness 1 mth
- Short- and long-term govt. securities rates, policy variable rate: Coverage Yes; GDDS periodicity M; Pakistan periodicity D (TBills) / M (other); GDDS timeliness In high-frequency publication; Pakistan timeliness Daily
- Money or interbank market rates and a set of deposit and lending rates: Coverage Yes; GDDS periodicity M; Pakistan periodicity M; GDDS timeliness Same as above; Pakistan timeliness Daily
- Share price index, as relevant: Coverage Yes; GDDS periodicity M; Pakistan periodicity D/W; GDDS timeliness Same as above; Pakistan timeliness One day
- Balance of payments aggregates: Coverage Yes; GDDS periodicity A; Pakistan periodicity Q; GDDS timeliness 6 mths; Pakistan timeliness 2 mths
- Public and publicly guaranteed external debt outstanding, with maturity breakdown: Coverage Yes; GDDS periodicity Q; Pakistan periodicity Q; GDDS timeliness 1-2 qtrs; Pakistan timeliness Provisional: 8-10 wks Revised: 6 to 8 additional wks
- Public and publicly guaranteed external debt service schedule: Coverage Yes; GDDS periodicity Twice yearly; Pakistan periodicity Q; GDDS timeliness 3-6 mths; Pakistan timeliness Same as above
- Private external debt not publicly guaranteed: Coverage Yes; GDDS periodicity A; Pakistan periodicity Q; GDDS timeliness 6-9 mths; Pakistan timeliness Same as above
- Gross official reserves denominated in U.S. dollars: Coverage Yes; GDDS periodicity M; Pakistan periodicity W; GDDS timeliness 1-4 wks; Pakistan timeliness 1 wk
- Reserve related liabilities: Coverage Yes; GDDS periodicity M; Pakistan periodicity M; GDDS timeliness 1-4 wks; Pakistan timeliness 4 wks
- Total exports and imports in U.S. dollars: Coverage Yes; GDDS periodicity M; Pakistan periodicity M; GDDS timeliness 8-12 wks; Pakistan timeliness 5 days
- Major commodity breakdowns with longer time lapse: Coverage Yes; GDDS periodicity M; Pakistan periodicity M; GDDS timeliness 8-12 wks; Pakistan timeliness Same as above
- Exchange rates: spot rates: Coverage Yes; GDDS periodicity D; Pakistan periodicity D; GDDS timeliness In high-frequency publication; Pakistan timeliness Daily

(Italics indicate encouraged categories.)

### SDDS assessment and issues to address prior to subscription
- Pakistan has expressed interest in subscribing to the SDDS in the near future; current practices were reviewed against SDDS requirements.
- Real sector issues:
  - All data categories meet SDDS coverage except annual data on wages/earnings, which are not disseminated.
  - Periodicity and timeliness shortcomings:
    - The first GDP releases: published about two weeks before the end of the fiscal year, based on nine to ten months’ observed data and two to three months’ projections. The first annual GDP data based on 12 month’s observations are disseminated with a lag of 11 months (vis-à-vis the recommended 6-9 months). This would be remedied with the publication of quarterly national accounts presently in preparation.
    - Employment/unemployment data: compiled on an annual basis and disseminated with a lag of six months (vis-à-vis the recommendation for quarterly periodicity and timeliness).
    - Wages/earnings: SDDS requires quarterly periodicity and timeliness; Pakistan does not meet this.
- Fiscal sector issues:
  - General government operations basically meet SDDS periodicity and timeliness requirements.
  - Central government operations are compiled only on a quarterly basis and disseminated with a lag of two months (against the SDDS requirement of monthly periodicity and timeliness).
- Financial sector issues:
  - Analytical accounts of the banking system, interest rates, and the share price index meet SDDS requirements.
  - Analytical accounts of the SBP do not distinguish the main aggregates clearly.
  - SBP data were being disseminated four weeks after the end of the reference month, against the required two weeks. Beginning in June 2004, the SBP disseminates the analytical accounts based on month-end data with a lag of two weeks.
- External sector issues:
  - Most data categories (balance of payments, international reserves and foreign currency liquidity, merchandise trade, and exchange rate) meet SDDS coverage, periodicity, and timeliness.
  - Only partial and non-disseminated quarterly data are available on the international investment position (vis-à-vis the annual periodicity and semiannual timeliness required by the SDDS).

### Data Quality Assessment Framework (DQAF) application and summary assessment
- Purpose and scope:
  - The IMF’s Statistics Department (STA) developed the DQAF based on international best practice to assess data quality across five dimensions: integrity, methodological soundness, accuracy and reliability, serviceability, and accessibility — plus prerequisites.
  - An assessment was conducted for six macroeconomic datasets: national accounts, consumer price index, wholesale price index, government finance, monetary, and balance of payments statistics, using dataset-specific DQAF frameworks.
- Overall conclusions:
  - Pakistan’s macroeconomic statistics and statistical base are considered broadly adequate by the IMF’s Middle East and Central Asia Department (MCD) for effective surveillance.
  - The mission identified shortcomings in some statistical practices that may detract from accurate and timely analysis of economic and financial policies.
- Prerequisites of quality — legal and institutional environment, resources, and quality awareness:
  - Legal/institutional:
    - General Statistics Act No. 69 of 1975 establishes the legal and institutional environment for compiling official statistics.
    - The Act does not explicitly state that the FBS is responsible for national accounts or price statistics, but this is implied by tradition and the FBS’s official status.
    - The Act guarantees confidentiality and specifies penalties for breaches; it provides the FBS legal basis to collect data from all national units, but enterprises’ response rates are problematic. Fines for noncompliance are not indexed and have become insignificant for many nonrespondents.
    - The National Statistical Council (NSC) coordinates statistical programs; committees on national accounts and price statistics (data providers and users) meet regularly.
  - Resources:
    - National accounts statistics: resources generally below desired levels; number of staff in the National Accounts Wing (NAW) broadly adequate, but technical skills may be insufficient; information technology resources are deficient and many computers are obsolete.
    - Price statistics: staff resources broadly adequate; Karachi office computer resources just sufficient; only a few field offices have computers to transmit price information by electronic mail; many offices transmit by fax or express mail.
    - Government finance statistics (GFS): MOF staff and computer resources adequate for limited GFS production but would benefit from further training; AGPR main constraint is lack of computer automation of GFS processing.
    - SBP: financial and computing resources allocated to statistical programs are adequate and used efficiently; human resources generally adequate though monetary statistics compilers require further training.
  - Quality awareness:
    - FBS management convenes regular senior staff meetings in NAW; committees with data providers and users meet regularly to review techniques and quality.
    - MOF: quality awareness should be strengthened; GFS compilers lack a user perspective and focus on processes rather than products. Planned improvements to budget accounting standards and systems exist but no specific processes to deal with quality considerations.
    - SBP: demonstrates quality awareness through published “Vision Statement” and Statistics Department’s business plan; Planning and Development Division addresses quality issues and provides guidelines on new and emerging data requirements.
- Integrity
  - FBS:
    - Recruitment and promotion follow public service regulations encouraging professionalism. Staff are appointed on a permanent basis after civil service entry requirements.
    - The head of FBS must be appointed from staff ranks and can remain until retirement; delays in appointing a Director-General have been a concern.
    - Decision on dissemination of national accounts rests with the FBS. The National Accounts Committee (chaired by the Secretary of the Statistics Division) reviews and approves accounts and may recommend revisions prior to dissemination.
    - FBS monitors press coverage and provides clarifications for erroneous interpretations.
    - Laws governing statistics collection, compilation, and dissemination are posted on the FBS website and in hard copy. Internal government access to some statistics prior to release is made known on the DSBB and in publications.
    - Staff behavior guided by civil service rules; ethical adherence promoted in training and annual performance reviews.
  - MOF/GFS:
    - MOF statistical policies guided by professional principles. AGPR has autonomy under law for accounting functions; longstanding convention of MOF independence in compiling and publishing fiscal data.
    - No evidence of political influence in choice of sources or methods. GFS widely accessible; disseminated on MOF website and in Economic Survey.
    - Limited information provided to users on production and dissemination arrangements. GFS compilers governed by Pakistan civil service Code of Conduct.
  - SBP:
    - Practices encourage professionalism; data compiled on an impartial basis; choices of sources and techniques based solely on statistical considerations.
    - SBP provides comments on erroneous interpretation and misuse; terms and conditions for data collection, processing, and dissemination are public; no internal government access prior to publication.
    - SBP products are identified as such and SBP requests attribution when statistics are used. Major methodological changes are noted at introduction and data suppliers are notified in advance.
    - SBP Staff Regulations provide guidelines for ethical standards.
- Methodological soundness
  - The DQAF assesses application of international standards, guidelines, and agreed practices across dataset-specific elements (concepts and definitions, scope, classification and sectorization, and basis for recording). The source begins the methodological soundness section but the detailed dataset-specific findings appear in the accompanying document.

*Source: _cr04398 - 17.      Regarding the external sector, the SBP plans to (i) continue the full implementation*

### 29.      The national accounts follow the concepts and definitions

### _cr04398 - 29.      The national accounts follow the concepts and definitions

### Concepts and definitions — national accounts, CPI/WPI, GFS, monetary statistics, balance of payments
- National accounts follow the System of National Accounts 1968 (1968 SNA).
- Scope of the annual accounts is very narrow and does not meet the minimum requirements for implementing the 1993 SNA.
- FBS is instituting a comprehensive plan to revise estimates and implement the recommendations of the 1993 SNA and expects to disseminate revised estimates, along with a set of quarterly accounts, during the first half of 2004.
- Current measurement of output:
  - Does not cover all own-account production and most informal activities.
  - Classification systems are broadly consistent with internationally accepted guidelines except household consumption, which is not classified by purpose.
  - Basis for recording is not fully in accordance with the 1993 SNA: sales taxes are excluded from the valuation of intermediate consumption.
  - Recording is on an accrual basis except for government transactions, which are recorded on a cash basis.
- CPI and WPI concepts and definitions follow international guidelines, with limitations:
  - CPI scope limited to 52 urban centers (covering about 68 percent of all households), excluding rural households.
  - WPI does not cover small-scale manufacturing enterprises.
  - National classification for household consumption differs significantly from international standards; national activity classification is close to international recommendations.
  - Goods and services are recorded at market values and on an accrual basis.
- GFS compilation:
  - Broadly based on A Manual on Government Finance Statistics, 1986 (GFSM 1986); treatment of privatization proceeds resembles GFSM 2001.
  - Definition of government debt includes government guaranteed debt (not fully following international standards).
  - No specific plans to migrate to GFSM 2001.
  - Scope covers budgetary general government but excludes extra-budgetary units in central government.
  - Classification and sectorization follow GFSM 1986 to a limited extent; institutional sectors are not explicitly defined.
  - Local governments are included with provincial governments, despite ongoing decentralization initiatives.
  - Foreign grants to provincial governments and loans received by provincial governments from the central government are incorrectly classified (as financing and revenue, respectively).
  - Expenditure classification departs substantially from GFSM 1986 methodology; economic and functional classifications are not clearly distinguished and level of detail is insufficient.
  - Basis of recording is on, or close to, a cash basis; transactions recorded on a gross basis; corrective transactions are not necessarily made in the original period.
- Monetary statistics:
  - Analytical framework reflects IMF’s draft Guide to Money and Banking Statistics in International Finance Statistics of December 1984.
  - Scope is comprehensive.
  - Classification and sectorization in the Statistical Bulletin do not follow international standards (e.g., positions with nonresidents not always identified; securities repurchase agreements not treated as collateralized loans; financial derivatives recorded off-balance-sheet).
  - Basis for recording flows and stocks largely consistent with MFSM, with accrued interest identified in the balance sheet but not included in the outstanding amount of the corresponding financial instrument.
  - SBP prepares additional data for internal use and IMF reporting that are broadly in line with the Monetary and Financial Statistics Manual; those data were not publicly available at the time of the mission and were not used for the assessment.
- Balance of payments:
  - Compiled broadly in accordance with BPM5 concepts and definitions, but presentation on the web and in SBP publications is still based on BPM4.
  - Scope regarding residency generally conforms to BPM5 with exceptions: enterprises in Export Processing Zones and offshore banks regarded as nonresidents; Pakistani nationals residing abroad but owning a dwelling in Pakistan considered residents.
  - Classification and sectorization follow BPM5 to some extent; specific misclassification example: entire amount of oil exports under product sharing agreements is classified as direct investment in the current account instead of allocating repatriation of direct investment capital to the financial account.
  - Basis for recording transactions follows best international practices.

### Accuracy and reliability — source data, statistical techniques, assessment and validation
- National accounts source data:
  - Derived primarily from benchmark censuses, a regular household budget survey, and administrative sources.
  - Comprehensive business register updated continuously; regular enterprise surveys are not conducted.
  - Census of manufacturing conducted every five years; other censuses infrequent.
  - Censuses suffer from poor coverage and very low response rates.
  - Administrative data timely, but delays in survey fieldwork hinder timely compilation.
- Statistical techniques for national accounts are weak:
  - Estimates compiled at 1980/81 prices.
  - Reliance on benchmarks over five years old is widespread.
  - Volume indicators derived from highly aggregated price indices, some not reflecting underlying price changes.
  - Procedures to adjust data sources to improve coverage are deficient; techniques to derive exhaustive measures of GDP are not employed.
- Assessment and validation for national accounts:
  - Sampling errors and non-response are assessed.
  - Discrepancies in statistical outputs are not investigated despite under-coverage and non-measurement of informal activities.
  - FBS has conducted revision studies as part of rebasing and 1993 SNA implementation, developing experimental estimates for major aggregates using revised sources and techniques.
- CPI/WPI source data and techniques:
  - CPI weights derived from Family Budget Survey; WPI weights from a variety of sources, some unsatisfactory.
  - Price collection surveys are broad-based; source data timely and reasonably approximate required definitions and accounting rules.
  - Most statistical techniques sound except:
    - No adjustments to Family Budget Survey results to resolve underreporting by households of certain expenditures.
    - Price index for housing calculated on basis of smoothed construction costs, which does not reflect international standards and has large repercussions given the large weight of housing.
  - Assessment and validation of source and intermediate data and outputs are adequate; revision studies were conducted when changing the base year.
- GFS source data and techniques:
  - Source data reasonably approximate GFSM 1986 requirements; compiled directly from accounting/administrative data, obviating other statistical techniques.
  - Assessment and validation mainly by vetting input data by line agencies and benchmarking preliminary data to audited final data.
  - No systematic reconciliation to ensure consistency with other sources.
  - Revisions result from replacement of preliminary by audited data; no systematic studies of revisions, though benchmarking examines significant changes for explanation.
- Monetary statistics source data and techniques:
  - Main source: highly aggregated weekly statements from the Banking Supervision Department lacking required breakdowns.
  - Source data timely except SBP end-June fiscal year data available with a two to three month lag.
  - Weekly data entered manually (potential for errors); automated procedures exist to identify processing errors.
  - Assessment and validation adequate: comparisons between banks’ weekly and monthly statements; plausibility checks applied; revision-study findings used internally.
- Balance of payments source data and techniques:
  - Main sources: International Transactions Reporting System (ITRS), enterprise reports, administrative records of government and SBP.
  - Data broadly align with international definitions, classifications, and scope, and are timely.
  - Statistical techniques generally sound (example: methodology to calculate the cif/fob ratio).
  - Assessment and validation include comparisons between sources (ITRS vs government records or customs returns).
  - Revision studies are not carried out.

### Serviceability — relevance, timeliness, periodicity, consistency, revisions
- National accounts:
  - Relevance monitored through National Accounts Committee (NAC) with public/private representatives.
  - Timeliness does not meet GDDS recommendations; data are compiled annually (meets GDDS but not SDDS periodicity).
  - Statistics are consistent within dataset, over time, and with balance of payments and government finance statistics.
  - Revisions follow a predetermined schedule but this is not communicated to the public.
- CPI and WPI:
  - Relevance monitored via expert groups and occasional users’ surveys.
  - Timeliness and periodicity meet GDDS and SDDS.
  - Estimates internally consistent; aggregate index invariant to typology of aggregation.
  - Revisions follow a well-established schedule; base year changed every ten years; users informed of causes of revisions.
- GFS:
  - Fiscal tables are designed for fiscal policy monitoring and extensively used.
  - MOF monitors press and forum comments but receives little direct feedback from users; no systematic procedures to ascertain relevance to users’ needs.
  - Timeliness and periodicity meet GDDS but not SDDS requirements.
  - GFS internally consistent except discrepancy between deficit/surplus and financing data, larger than expected.
  - Annual GFS time series consistent over time as percentage of GDP.
  - GFS not reconcilable with balance of payments or monetary statistics.
  - Revisions usually follow a regular schedule; preliminary data clearly identified.
  - Corrections are made to data in the period discovered rather than shown as revisions to the original data; no published analysis or explanation of revisions.
- Monetary statistics:
  - Relevance monitored in interdepartmental SBP meetings; no formal external consultation process.
  - Analytical SBP and banking sector accounts meet GDDS timeliness and periodicity except SBP end-June data disseminated with a two-three months lag.
  - Large discrepancies exist between SBP and scheduled banks positions.
  - Consistency checks with balance of payments done regularly; not reconcilable with government finance statistics.
  - No formal revision policy; monetary data are final when first released and revisions are infrequent.
- Balance of payments:
  - Relevance monitored through the Balance of Payments Standing Committee; limited external user feedback.
  - Data exceed GDDS recommendations and meet SDDS requirements for periodicity and timeliness.
  - Data internally consistent and verified against monetary statistics, but not with government finance statistics.
  - Revisions follow a regular, well-established, and transparent cycle; preliminary data are identified as such.

### Accessibility — data and metadata dissemination, user assistance
- National accounts accessibility:
  - Presented clearly with varying levels of detail; brief analysis provided in national accounts publication.
  - Data not disseminated according to a preannounced schedule though users aware of stable timeframe.
  - Data presented to NAC before finalization and wider dissemination.
  - Metadata accessibility generally sound: FBS website, DSBB, and national accounts publication contain detailed information on sources and methods.
  - Assistance to users broadly adequate but could be improved by providing contact information for specific subjects in FBS publications and website.
  - Catalog of publications available in hardcopy and on the website.
- CPI/WPI accessibility:
  - Data and metadata mostly satisfactory; main publication facilitates interpretation and comparisons.
  - Family Budget Survey (source of CPI weights) has not been published.
  - Dissemination media include e-mail, website, and hard-copy publications.
  - Both indices released according to a preannounced schedule.
  - Hardcopies released to selected government users and press shortly before website publication.
  - Assistance broadly adequate; could be improved by providing contact information for specific subjects and publishing a catalog of FBS services.
- GFS accessibility:
  - GFS data have limited accessibility.
  - Dissemination media and format adequate; data released regularly but not on a preannounced schedule.
  - GFS tables reflect MOF internal needs and lack clear structure; no explanations or charts; no time series of sub-annual data.
  - Quarterly data presented on a cumulative basis, obscuring individual quarters.
  - GFS released to all users simultaneously; no non-published subaggregates.
  - GFS metadata accessible only on the DSBB in summary form.
  - Assistance to users inadequate: contact person publicized only on DSBB with telephone and facsimile numbers but not e-mail address.
- Monetary data accessibility:
  - Needs improvement; SBP balance sheet in the Statistical Bulletin not presented in an analytically useful format (e.g., no information on SBP’s foreign liabilities).
  - Statistics released simultaneously to all interested parties; nonconfidential unpublished data available upon request.
  - Metadata available on SBP website via hyperlink to DSBB.
  - Users provided contact information on DSBB and SBP website but not in hardcopy publications.
  - List of publications provided in the SBP’s Statistical Bulletin.

*Source: _cr04398 - 29.      The national accounts follow the concepts and definitions*

### 48.      The accessibility

### _cr04398 - 48.      The accessibility

### Accessibility of data and metadata
- The accessibility of data and metadata on balance of payments statistics is good.
- Statistics and metadata are presented in a clear manner that facilitates proper interpretation and meaningful comparisons in the Statistical Bulletin and in Pakistan’s Balance of Payments.
- These publications are posted on the SBP website, which is also hyperlinked to the DSBB.
- The data are released simultaneously to all users.
- Nonpublished but nonconfidential data are made available upon request.
- Users are provided with contact information on the website but not in hardcopy publications.
- A list of statistical publications is included in the Statistical Bulletin and Pakistan’s Balance of Payments.
- Note: In June 2004, the SBP began disseminating monetary statistics more in line with the MFSM. Also, a publicly announced advance release calendar is available on the SBP website.

### Staff recommendations — General (IV. STAFF RECOMMENDATIONS)
- Priority: The actions listed should be treated with high priority; those regarding national accounts and government finance statistics need to be addressed with the greatest urgency.
- A more detailed list of recommendations is included in the accompanying Detailed Assessments document.

General recommendations
- Pursue the initiative to give more autonomy to the FBS and conduct public awareness exercises aimed at improving the public perception of the FBS.
- Undertake actions to improve response rates for surveys, for example, by amending the General Statistics Act to increase the fines for not providing information.
- Undertake a review of statistical policies and practices in consultation with users, addressing at least the following issues:
  - (i) presentation of statistics (tables, text, charts, etc.);
  - (ii) development of detailed metadata for all datasets;
  - (iii) establishment of advanced release calendars;
  - (iv) simultaneous release of data to all users;
  - (v) publicizing the availability of unpublished data;
  - (v) conducting revision studies; and
  - (vii) assisting users of official statistics by providing more detailed contact information.
- Provide more training in statistical methodologies and modernize computer equipment.
- Strengthen coordination among institutions to enhance intersectoral consistency of data.
- Update and enhance regularly the GDDS metadata posted on the DSBB for all sectors.

### Staff recommendations — National accounts
- Amend the role and functions of the NAC so that the final decision on the dissemination of the statistics rests with the FBS.
- Introduce a program of regular (annual) establishment surveys to improve the timeliness of source data. Broaden the economic activities covered by surveys.
- Improve the timeliness of the national accounts. Disseminate the estimates that are revised for government agencies six months after the end of the reference period.
- Implement procedures to estimate the non-observed economy, targeting informal activities significant in Pakistan.

### Staff recommendations — Consumer and wholesale price indices
- Implement the planned introduction of the COICOP as soon as possible.
- Review the treatment of Housing in the CPI, preferably by replacing the technique for estimating rentals on the basis of construction costs with an index reflecting the cost of purchasing and maintaining dwellings.
- Consider expanding the scope of the CPI to cover all urban areas, and eventually cover all Pakistan by including information on rural areas.

### Staff recommendations — Government finance statistics
- Set up a small group within the MOF whose main task is the compilation and dissemination of GFS.
- Migrate to the GFSM 2001 classification structure, as far as possible.
- Provide the classificatory detail specified in the GFSM 2001 and publish a time series of discrete quarterly data.
- Examine the statistical discrepancy between the deficit/surplus (net lending/borrowing under GFSM 2001) and financing for each quarter, under the supervision of the Fiscal Monitoring Committee.

### Staff recommendations — Monetary statistics
- Compile and disseminate monetary statistics based on the data collected through the monthly report forms rather than the weekly statements provided by the Banking Supervision Department.
- Enhance the existing data validation procedures to ensure that mutual claims between the SBP and the scheduled banks are recorded consistently.
- Implement fully the guidelines of the MFSM on sectorization, classification, and accounting.
- Note: Beginning in June 2004, monetary statistics are based on month-end data and are more in line with the MFSM.

### Staff recommendations — Balance of payments statistics
- Compile data on the external transactions of enterprises operating in the Export Processing Zones and offshore banks and treat them as residents of Pakistan.
- Disaggregate the counterpart of exports of crude oil under product sharing agreements into:
  - (i) an income component, to be classified in the direct investment income account; and
  - (ii) a component for the repatriation of direct investment capital, to be classified as foreign direct investment in Pakistan in the financial account.
- Strengthen the procedure to estimate reinvested earnings to include, in addition to earnings capitalized as bonus shares, all earnings capitalized in other reserve accounts.
- Reclassify government-guaranteed loans from the general government sector to the sector of the beneficiary of the loan proceeds.
  - Note: This recommendation has been adopted in 2004.
- Complete the compilation of the international investment position on the basis of BPM5.

### Data Quality Assessment Framework — Summary presentation for Pakistan (Table 2 highlights)
- 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.
- Prerequisites of Quality
  - 0.1 Legal and institutional environment: NAS = LO; CPI = O; WPI = O; GFS = LO; MS = O; BOPS = O.
  - 0.2 Resources: NAS = LO; CPI = LO; WPI = LO; GFS = LO; MS = LO; BOPS = O.
  - 0.3 Quality awareness: NAS = O; CPI = O; WPI = O; GFS = LO; MS = O; BOPS = O.
  - Comment: NA, CPI, WPI (0.1): Authorities plan to propose legislation to give more autonomy to the FBS (short term). MS (0.2): SBP plans to increase resources dedicated to training in the monetary statistics methodology and compilation methods, including international statistical guidelines (short term).
- Integrity
  - 1.1 Professionalism: NAS = LO; CPI = O; WPI = O; GFS = O; MS = O; BOPS = O.
  - 1.2 Transparency: NAS = O; CPI = O; WPI = O; GFS = LO; MS = O; BOPS = O.
  - 1.3 Ethical standards: NAS = O; CPI = O; WPI = O; GFS = O; MS = O; BOPS = O.
- Methodological soundness
  - 2.1 Concepts and definitions: NAS = LO; CPI = O; WPI = O; GFS = LO; MS = LO; BOPS = LO.
  - 2.2 Scope: NAS = LNO; CPI = LO; WPI = O; GFS = LO; MS = O; BOPS = LO.
  - 2.3 Classification/Sectorization: NAS = LO; CPI = LO; WPI = O; GFS = LNO; MS = LNO; BOPS = LO.
  - 2.4 Basis for recording: NAS = LO; CPI = O; WPI = O; GFS = LO; MS = LO; BOPS = O.
  - Plans: BOP (2.1, 2.2, 2.3): SBP plans to adopt fully the format of the BPM5 for the presentation of data (short term). CPI & WPI (2.1,2.3): FBS plans to disseminate data on a producer price index, and to introduce COICOP for the CPI (short term). GFS (2.3): MOF has expressed its intention to adopt the GFSM 2001 (medium term). MS (2.1,2.3,2.4): SBP plans to fully adopt the MFSM for the compilation of monetary statistics (medium term).
- Accuracy and Reliability (selected elements)
  - 3.1 Source data: NAS = LNO; CPI = O; WPI = LO; GFS = O; MS = LNO; BOPS = O.
  - 3.2 Statistical techniques: NAS = LNO; CPI = LO; WPI = O; GFS = O; MS = LO; BOPS = O.
  - 3.3 Assessment and validation of source data: All sectors = O.
  - 3.5 Revision studies: NAS = O; CPI = O; WPI = O; GFS = LO; MS = O; BOPS = LNO.
  - Plans: NA (3.1,3.2): FBS plans to extend surveys to improve data sources, and change the base year to 1999/2000 (short term). GFS (3.1): MOF plans to implement a revised chart of accounts and a new accounting model compatible with accrual accounting and reporting (medium term). BOP (3.5): SBP plans to conduct studies of revisions (short term).
- Serviceability (selected elements)
  - 4.1 Relevance: NAS = O; CPI = O; WPI = O; GFS = LO; MS = LO; BOPS = LO.
  - 4.2 Timeliness and periodicity: NAS = LO; CPI = O; WPI = O; GFS = O; MS = LO; BOPS = O.
  - 4.3 Consistency: NAS = O; CPI = O; WPI = O; GFS = LO; MS = LNO; BOPS = LO.
  - 4.4 Revision policy and practice: NAS = LO; CPI = O; WPI = O; GFS = LO; MS = LO; BOPS = O.
  - Plans: NA (4.2): FBS expects to disseminate quarterly GDP data by activity and category of expenditure in early 2004. GFS (4.2): MOF plans to disseminate monthly GFS with a lag of one month (short term). MS (4.1, 4.2, 4.3, 4.4): SBP plans to improve data consistency between the central bank and scheduled banks, and between MS and GFS (short term); and inform users of its revision policy and practice and include documentation of revisions in the statistical publications (short term). BOP (4.1): SBP plans to fully adopt the presentation format of BPM5 (short term).
- Accessibility (selected elements)
  - 5.1 Data accessibility: NAS = LO; CPI = LO; WPI = LO; GFS = LO; MS = LO; BOPS = O.
  - 5.2 Metadata accessibility: NAS = O; CPI = LO; WPI = O; GFS = LNO; MS = LO; BOPS = O.
  - 5.3 Assistance to users: NAS = LO; CPI = LO; WPI = LO; GFS = LNO; MS = LO; BOPS = LO.
  - Plans: GFS (5.2): MOF plans to develop and disseminate metadata (short term). MS (5.1,5.2): SBP plans to continue to improve the presentation of monetary statistics to make them fully in accordance with the MFSM, and also to disseminate more detailed metadata.

### Authorities' response — Overview and selected actions (Response by the Authorities, November 29, 2004)
- Overview
  - The authorities (FBS, MOF, SBP) appreciate the IMF’s data module ROSC report and welcome its recommendations to improve Pakistan’s statistical database and meet SDDS requirements.
- FBS — General recommendations and actions
  - Recommendation: Pursue autonomy for FBS via merger of statistical bodies.
    - Action/Status: A proposal for creation of a single, fully professional, dynamic and autonomous statistical organization through merger of Technical Wing of Statistics Division, FBS, PCO and ACO has been submitted to the Government and is being pursued; expected to be completed during 2005.
  - Public awareness: Data users—producers meetings, workshops, press briefings, website of the Statistics Division; meetings with private sector Chambers of Commerce and Industries and associations.
  - Recommendation: Increase fines under the General Statistics Act, 1975 to improve survey response.
    - FBS view: Does not consider increased fines will work; seeks help of CBR and provincial departments to improve response rates for Large Scale Manufacturing Industries; response rate for Household Surveys is quite encouraging.
  - Recommendation: Review statistical policies/practices; establish advanced release calendar; simultaneous release; provide contact information.
    - Action/Status: NSC/FSC reviews; advance release calendar on FBS website; data simultaneously released and posted; contact information (names, e-mail, phone, fax, postal address) provided on publications and website.
  - Recommendation: Increase staff training and modernize computer equipment.
    - Action/Status: Project proposal on Capacity Building of the Statistics Division with World Bank sent to MOF; proposal for assistance to Training Wing negotiated with UNFPA.
  - Recommendation: Strengthen coordination among institutions.
    - Action/Status: Regular meetings between producers and users planned.
  - Recommendation: Update GDDS metadata on DSBB regularly.
    - Action/Status: Updates are being done regularly by the FBS.
- FBS — National accounts specific responses and plans
  - FBS does not support increasing fines per General recommendation 2 (reiterated).
  - Plans to improve computing resources: replace outdated 386/486 PCs, network offices, seek German support for Regional/field Offices and National Accounts.
  - Training: IMF and UNFPA assistance welcomed; IMF may provide training on seasonality estimation and smoothing of quarterly national accounts.
  - NAC role: FBS states NAC is a technical advisory body and does not create external influence on statistics.
  - Vacancy of Director General: Reorganization will place heads of organs; stakeholders are informed; no uncertainty.
  - Supply and use tables and institutional sector accounts: FBS will develop both on annual basis; capacity requirements being worked out.
  - Coverage of merchandise trade and unrecorded cross-border trade: No mechanism currently exists; IMF technical assistance welcomed.
  - Non-observed economy: IMF technical assistance welcomed.
  - Rebasing: FBS agrees with five-year rebasing cycle; next rebasing will be for year 2005-06 after which a five year cycle will be observed; requires additional financial and manpower resources.
  - Private final consumption from HIES: Agreed; will be done every other year as HIES/Living Standards data available on alternate years.
  - Timeliness: Efforts underway to prepare and release quarterly national accounts.
  - Quarterly surveys: Employment/unemployment
    - Current practice: LFS annual with sample size of about 21,000 households, evenly distributed among four quarters.
    - Plan to produce quarterly estimates: Increase sample size to 32,000 households (8,000 households for each quarter) to make quarterly estimates representative with rural-urban breakup; this will entail additional funds and manpower; FBS plans to conduct survey on quarterly basis through its regular resources in 2005-06.
  - Wage rates: Currently no Wage Rate Survey; IMF technical assistance welcome.

*Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04398.pdf*

### 3. Consumer and Wholesale Price Indices

### 3. Consumer and Wholesale Price Indices

### Classification and index basis
- Recommendation 1: The National classifications are used instead of the internationally recommended ones. Implement the planned introduction of the COIPOP as soon as possible.
- Response: "This is a good recommendation and is under consideration."

### Family Budget Survey and CPI weights
- Recommendation 2: The Family Budget Survey, which is the source of the weights for the CPI, has not been published. No adjustments are made to the results from the Family Budget Survey to resolve under reporting by households of certain types of expenditures.
- Response and factual details:
  - "Family Budget Survey was conducted in 2000-01."
  - "Its results are not published as the same are used only for change of base of CPI and SPI."
  - "Under reporting of expenditures was taken care of to the extent possible at the editing/coding stage of the Family Budget Survey."
  - "However, next survey will be published."

### Treatment of housing in the CPI
- Recommendation 3: Review the treatment of Housing in the CPI, preferably by replacing the technique for estimating rentals on the basis of construction costs with an index reflecting the cost of purchasing and maintaining dwellings.
- Response: "This is a good suggestion and we would welcome technical assistance from the IMF to handle house rent in CPI."

### CPI geographic scope and coverage
- Recommendation 4: The CPI scope is limited to 52 urban centers (covering about 68 percent of all urban households), thus excluding households living in rural areas. Consider expanding the scope of the CPI to cover all urban areas (as allowed by the 2000/2001 Family Budget Survey), and eventually cover all Pakistan by including information on rural areas.
- Response and factual clarifications:
  - "The CPI scope extends to 52 urban centers covering 68 percent of urban population (not 68 percent of households)."
  - "Organized markets do not exist in the rural areas."
  - "Rural people purchase goods and services for consumption from nearby towns/small cities. Hence, small cities were covered in CPI."

### Coverage of the Wholesale Price Index (WPI)
- Recommendation 5: WPI does not cover small scale manufacturing enterprises.
- Response: "WPI covers large-scale as well as small-scale manufacturing enterprise."

*Source: _cr04398 - 3. Consumer and Wholesale Price Indices (excerpt)._

### 0.1 Legal and institutional environment

### 0.1 Legal and institutional environment

### Legal framework and responsibilities
- Two laws establish the legal framework for official statistics: the Industrial Statistics Act No. 19 of 1942 and the General Statistics Act No. 69 of 1975.
- Industrial Statistics Act No. 19 of 1942:
  - Limited in scope; covers activities of the provincial governments in collecting data from business establishments.
  - Used as basis for conducting the census of manufacturing industries (CMI), undertaken by provincial governments.
- General Statistics Act No. 69 of 1975:
  - Establishes broad responsibility at federal and provincial levels for collecting, processing, and disseminating statistics.
  - Article 5 prescribes appointment of a Federal Statistics Authority and provincial statistics authorities.
  - The Secretary of Statistics is the Federal Statistics Authority and head of the central Statistics Division (SD).
  - The SD comprises three attached departments: the Federal Bureau of Statistics (FBS), the Population Census Organization, and the Agricultural Census Organization.
  - Tasks of census organizations are limited to their census activities; the FBS is responsible for compiling general statistics on all activities in Pakistan.
  - While the General Statistics Law does not explicitly state that the FBS is responsible for compiling national accounts statistics, this responsibility is implied and has been accepted nationally and internationally.

### Data sharing, coordination, and organizational network
- Article 3 of the General Statistics Act provides for formation of a National Statistical Council to coordinate activities; the Council has not been active in practice.
- Much coordination has been undertaken by the National Accounts Committee (NAC), which comprises representatives from most data-producing agencies.
- FBS reports no difficulty receiving administrative data from other agencies.
- FBS organizational network:
  - 16 regional offices
  - 19 field offices (one level below regional offices)
  - 131 statistical cells operating out of various ministries and government agencies (e.g., cells at customs stations compile trade data)
- Census of Manufacturing Industries (CMI):
  - Conducted every five years as joint effort between provincial offices and the FBS; provincial offices operate independently but agree on timing and execution.

### Confidentiality and legal protections
- Confidentiality protected under both the General Statistics Act of 1975 and the Industrial Statistics Act of 1942.
- General Statistics Act:
  - Article 11: information collected shall not be used for any purpose other than compilation of statistics; shall not be published in a form that discloses the state of affairs of the unit; disclosure only with written consent of respondent.
  - Information cannot be used as evidence against respondent unless prosecuting for contravening provisions under the Act.
  - Article 15: establishes penalties for breach of confidentiality (fine or imprisonment or both); amended by Ordinance No. 7 of 2000 to increase the fine from 1,000 rupees to 7,000 rupees.
- Industrial Statistics Act:
  - Article 7 provides safeguards; Article 9 stipulates penalties for improper disclosure.
- User concerns:
  - IMF user survey: many respondents expressed reservations that FBS might transfer data to Central Board of Revenue (CBR) for assessments; no evidence found of breaches.
- FBS measures to protect confidentiality:
  - Provincial surveys processed provincially and submitted to federal office in aggregate form or after removal of identifying records.
  - Federal-level data access restricted to individuals engaged in processing.
  - Questionnaires and confidential material destroyed by incineration per civil service administrative regulations.

### Legal mandate and enforcement of reporting
- Article 9 of the General Statistics Act authorizes collection of any information deemed relevant; individuals are obligated to provide information.
- Article 10: after due notice, statistical agents may access statistical records or premises believed to hold relevant records.
- Article 12: penalties for noncompliance:
  - Fine for noncompliance is 500 rupees.
  - For establishments, a further fine of up to 200 rupees for each day the order is violated.
- Notes on enforcement:
  - Fines are very low and have not been increased since 1975.
  - Despite high non-response, penalties are rarely employed; process for pursuing nonrespondents largely ineffective.
  - Nonresponse often cites confidentiality concerns (possible sharing with CBR).

### Resources (staff, financial, computing)
- National Accounts Wing (NAW) staffing:
  - 61 staff positions with 10 vacancies (mainly clerical level).
  - Staff require training in national accounts statistics.
  - Staff turnover is low; FBS does not foresee an increase in turnover rates over the next five years.
- Financial resources:
  - Recurrent activity funding broadly adequate, but delays in disbursement by MOF have hindered task completion.
- Equipment and transportation:
  - Inadequate resources; critical deficiency in computing resources.
  - NAW lacks a central processing facility and is not connected to the FBS data processing unit.
  - Number of computers in NAW is below required; available machines are obsolete (primarily 386 and 486 processors), unable to support modern software desired for compilation.
  - Inefficient GFS consolidation practice is directly associated with paucity of computing resources: NAW staff compile GFS tables using hardcopy budget statements despite data being available electronically to MOF.
- Resource efficiency measures:
  - FBS ensures similar definitions and classifications across federal and regional levels to reduce processing errors.
  - FBS prepares a surveys plan in May for the upcoming financial year to schedule surveys efficiently and minimize need for additional temporary staff.
  - Survey instruments reviewed annually to minimize questionnaire size; surveys pooled where possible to reduce processing costs.
  - Department heads meet monthly to review expenditure and allocate funds for upcoming month.

### Quality awareness and quality monitoring
- Quality commitment:
  - FBS mission statement commits to providing accurate, reliable, timely, comprehensive, and user-friendly information.
  - Pakistan’s participation in IMF’s GDDS and posting metadata on the DSBB illustrate quality awareness.
- National Accounts Committee (NAC):
  - Formed by Government of Pakistan Notification No. 71(6)EEP/68 Islamabad, on November 5, 1968.
  - Functions:
    - to examine and assess data requirements for national accounts and determine priorities for collecting comprehensive data;
    - to review FBS conceptual approach and recommend improvements;
    - to examine and approve national accounts for release; and
    - to recommend research in national accounts.
  - Chaired by the Secretary of the SD; Director-General of FBS serves as secretary.
  - Membership: heads of public sector agencies including the State Bank of Pakistan (SBP), directors of provincial statistical offices, and representatives from private sector organizations.
  - NAC meets at least twice annually to review NAW activities; FBS staff prepare working papers for NAC consideration.
  - FBS does not conduct regular formal user surveys, considering NAC representative of main users.
- Trade-offs and planning:
  - FBS awareness of timeliness vs. accuracy trade-off is shown by schedule for compiling and revising statistics:
    - FBS compiles a first set of estimates principally to meet needs of national planners and later produces a revised set using comprehensive data.
    - Note: FBS produces a set of provisional estimates in May for MOF planners to prepare economic review for coming financial year beginning in July.
  - Training: Seminars on national accounts and computing techniques conducted regularly by FBS Training Wing.

### Integrity and professionalism
- Appointments and conditions:
  - Terms and conditions for FBS staff follow Civil Servants Act No. 71 of 1973.
  - Director-General appointed by the Prime Minister upon recommendation of the Establishment Division and the SD; must be selected from agency ranks; removable only on recommendation for an offense in public service rules.
  - Staff selected from public service recruits who meet entry requirements and pass entrance examinations.
- Management vacancy:
  - The post of Director-General is currently vacant following retirement in May 2003; delays in appointment have caused concern among users.
- Independence in compilation:
  - Choices of sources and techniques are informed solely by statistical considerations.
  - FBS conducting first comprehensive revision of compilation process since 1980/81; compiling case studies and proposing techniques based on good practices and data-collection constraints.
- Right to correct misuse:
  - FBS is entitled to respond to and does respond to misuse of statistics in national press, verifying and requesting corrections where necessary.

### Transparency and dissemination practices
- Terms and conditions public:
  - Collection and processing terms are detailed in the General Statistics Act; relevant Articles summarized or quoted on survey instruments and in correspondence.
  - Terms and conditions are available as part of GDDS metadata.
- Advance access:
  - Ministries of Finance and Planning receive national accounts statistics before public release; identified in FBS Brochure, 2002-2003 and posted on DSBB.
- Product identification:
  - All FBS publications identify the agency as producer; users are asked to acknowledge source of data in research and articles.
- Advance notice of methodological changes:
  - FBS will implement comprehensive changes to national accounts methodology and compilation practices during months following date of assessment.
  - FBS has kept users informed via website, annual national accounts publication, and FBS brochure.

### Ethical standards and staff conduct
- Staff guidelines:
  - Staff behavior guided by civil service rules compiled in Civil Service Establishment Code, made available upon hiring.
  - Annual performance appraisals assess staff conduct.
  - Training sessions for survey fieldwork remind staff to adhere to guidelines of good behavior.

### Methodological soundness: concepts, classifications, and scope
- Concepts and standards:
  - National accounts compiled according to System of National Accounts 1968 (1968 SNA).
  - FBS instituting comprehensive revision plan to implement recommendations of System of National Accounts 1993 (1993 SNA).
  - FBS expects to disseminate revised estimates, along with a set of quarterly accounts, during 2004.
- Accounts compiled (part of ISWGNA minimum requirements for 1993 SNA implementation):
  - annual value added and GDP at current and constant prices by activity;
  - annual expenditures on GDP at current and constant prices; and
  - annual value added components at current prices by activity.
- Gaps relative to ISWGNA minimum requirements:
  - FBS does not fulfill full ISWGNA minimum requirements:
    - Experimental institutional sector accounts compiled for 1984/85, 1989/90, and 1990/91 only (ISWGNA requires annual compilation).
    - FBS does not disseminate quarterly GDP estimates.
    - FBS does not disseminate annual supply and use tables as recommended.
- Coverage:
  - Accounts cover most of Pakistan’s economic territory, including territorial enclaves in the rest of the world (e.g., embassies) and workers who work part of the year in another country.
  - Exclusions from scope of Household Income and Expenditure Survey (HIES):
    - the Federally Administered Tribal Areas; and
    - military restricted areas.

### Production and assets boundaries (differences from 1993 SNA)
- Production boundary: not fully in accordance with 1993 SNA.
  - Included in output measurement:
    - mineral exploration;
    - production of computer software; and
    - production of entertainment, literary or artistic originals.
  - Not measured:
    - own-account production of goods for final consumption;
    - research and development on own-account;
    - output of goods for own account fixed capital formation; and
    - illegal output sold to willing buyers.
- Assets boundary: not in accordance with 1993 SNA.
  - Tangible assets not included:
    - defense related assets that could be used for civilian purposes;
    - valuables and historical monuments; and
    - agricultural work in progress.
  - Intangible assets not included:
    - mineral exploration;
    - Systems and standard applications computer software and databases;
    - entertainment, literary, and artistic originals;
    - patented entities; and
    - leases and other transferable contracts.

*Source: _cr04398 - 0.1 Legal and institutional environment*

### 2.3       Classification/sectorization

### _cr04398 - 2.3       Classification/sectorization

### Classification and sectorization systems
- The FBS has developed:
  - the Pakistan Standard Industrial Classification to classify activities, consistent with the second revision of the International Standard Industrial Classification.
  - the Pakistan International Product Classification to classify products, consistent with the Central Product Classification.
- Current expenditure of general government is classified using the Classification of the Functions of Government.
- Annual national accounts statistics are not classified by institutional sector.
- Household final consumption expenditure is not classified by purpose; no functional classification of consumption such as the Classification of Individual Consumption by Purpose is used.

### Basis for recording: valuation, timing, and grossing/netting
- Valuation rules partially follow the 1993 SNA:
  - Total imports and exports are valued on an f.o.b. basis.
  - Information on insurance and freight is available.
- Deviations from 1993 SNA recommendations:
  - Market output is valued at factor cost and not at basic or producer prices.
  - Intermediate consumption excludes sales tax.
- The national accounts are compiled according to the fiscal year of the government: July to June.
- Recording basis:
  - Transactions are recorded on an accrual basis, except for government transactions, which are recorded on a cash basis.
  - Work-in-progress for construction is recorded in the period it is produced.
  - Agriculture work-in-progress is not estimated; production of crops is recorded in the period of harvest. Revised estimates based on the new base year will include estimates for agriculture work-in-progress.
- Grossing/netting:
  - Transactions between establishments of the same enterprise are recorded on a net basis.

### Source data — coverage, timeliness, and main source programs
- Business register and enterprise coverage:
  - The FBS maintains a business register used as the sample frame for censuses and surveys of enterprises; the register is updated on a continuous basis.
  - The register contains about 56,000 records of corporations and includes information such as number of employees, major type of activity, and start-up capital.
  - The register may not represent the number and range of enterprises operating in Pakistan and may need to be updated with a business census.
- Use of surveys and administrative sources:
  - The FBS does not conduct regular enterprise surveys; current data are derived primarily from administrative sources.
  - A Pakistan Integrated Household Survey, incorporating an HIES component, was last conducted in 2000/01.
  - A CMI is conducted at regular five-year intervals.
- Census of Manufacturing Industries (CMI):
  - CMI is collaborative: FBS coordinates, provincial governments undertake fieldwork, provincial statistical offices process responses.
  - The census is based on factories lists maintained under the Factories Act of 1934; lists are not updated regularly and may not represent the manufacturing industry.
  - The response rate for the 2000/01 census was 46 percent of 9,800 factories surveyed.
  - Delays and low resources in provincial offices limit timeliness and quality. FBS has little control over execution.
  - FBS developed a quarterly survey of large-scale manufacturing to improve coverage; will use an economic census–derived frame replacing the current business register frame.
- Household Income and Expenditure Survey (HIES) 2000/01:
  - Covered all urban and rural areas but excluded military restricted areas and protected areas of the North-West Frontier Province (about 3 percent of the total population).
  - Urban frame developed by quick count listing; rural frame used list of villages from the 1998 Housing and Population Census.
  - Two-stage random sampling: PSUs selected with probability proportional to size; SSUs selected by systematic sampling.
  - Twelve households selected per urban PSU and 16 per rural PSU.
  - Total selected households: 14,912 from 1,050 PSUs.
  - Sixty-two percent of households selected were from urban areas.
  - Response rate: high, with only 233 refusals after two rounds of visits.
  - Data from the HIES are not widely used in the estimation process.
- Administrative data:
  - Government operations data sources:
    - Federal – MOF
    - Provincial – Provincial finance departments
    - District – District governments
    - Towns – Town committees
  - Budgetary data are received in hard copy and processed by the FBS; detailed data are available to measure output, intermediate consumption, fixed capital formation, and final consumption expenditure of government services. Defense-related expenditures are not available by type.
  - Balance of payments statistics are received from the State Bank of Pakistan.
  - Detailed foreign trade statistics are compiled by the FBS and available within four weeks of the end of the period but cover only trade documented by customs authorities. Unrecorded cross-border trade is not recorded and no estimates are made.
- Ad hoc surveys and censuses conducted (examples and years):
  - Census of Electricity Undertakings – 1999/2000
  - Census of Mining Industries – 1999/2000
  - Economic Census – 2000
  - Rent Survey of Dwellings – 1998/99
  - Survey of Inland Commercial Fishing – 1999/2000
  - Census of Software Industry and Related Services – 2002
- Assessment of reliance on censuses:
  - Reliance on censuses is costly and time-consuming; high cost has reduced usefulness to national accounts and resource constraints have compromised fieldwork quality.
- Source data quality and timeliness:
  - Administrative source data reasonably approximate definitions, scope, valuation, and time of recording required.
  - Definitions and classifications used in the HIES and CMI are consistent with national accounts methodology.
  - Administrative source data are timely; survey source data timeliness has been affected by delays in fieldwork completion.

### Statistical techniques — compilation and deflation methods
- Production approach:
  - Estimates of gross output and intermediate consumption are compiled largely at the two-digit level of the PSIC.
  - The use of fixed ratios is widespread; many are derived from benchmarks more than five years old.
  - Many estimation methods assume a fixed relationship in volume terms between intermediate consumption and output since the benchmark year.
- Summary table (sources and deflation/reflation methods) — methodology by activity (verbatim):
  - Agriculture: Current data on output volumes of crops obtained from provincial governments. Reflated using the wholesale price index.
  - Livestock: Intercensal growth rates of livestock censuses for 1976, 1988, and 1996. Reflated using the WPI.
  - Large-scale manufacturing: 1) CMI – used when data are available. 2) Quantum index of manufacturing to be applied to base year estimates of value added. Reflated using the relevant section of the WPI.
  - Small-scale manufacturing: Growth rate of the volume of small-scale manufacturing obtained through surveys conducted in 1987/88 and 1996/97. Reflated using relevant section of WPI.
  - Construction: Data on coefficient of value added to GFCF obtained from government agencies, including the Public Works Department, and construction companies. Coefficient applied to GFCF estimates. Value added deflated using wage index of construction workers.
  - Electricity and gas distribution: Data on volume and price of electricity and gas generated and distributed obtained from the producers. Deflated using unit value indices of electricity and gas.
  - Transport, storage and communication: 1) Data on railways, air transport, water transport, and communications obtained from relevant enterprises. 2) Data on mechanized road transport obtained FBS survey in major cities. 3) Data on non-mechanized transport obtained from town committees. 1) Railways deflated using unit value indices of passenger and tonnage kilometers for railways. 2) Air and sea transport deflated using passenger and tonnage kilometers for air and sea transport.
  - Wholesale and retail trade: 1) Data from trade statistics on imports and from industrial production statistics on domestic production of goods. 2) Data on marketed ratios and trade margins obtained from surveys of wholesale/retail conducted in 1984/85. Deflated using unit value indices of imports and price indices of domestic production.
  - Finance and insurance: Data obtained from State Bank of Pakistan and from other financial intermediaries. 1) Wages deflated using wage index of employees. 2) Other goods and services deflated using GDP implicit deflator.
  - Owner-occupied dwellings: 1) Data on number of dwellings derived from 1980/81 Population and Housing Census. 2) Rental values for urban areas from rental survey 1985. 3) Rental values in rural areas from 1984/85 HIES. 1) Reflation of volume estimates using number of dwellings.
  - Community, social, and personal services: Number of employees in service industry derived from 1980/81 census. CPI is used as a deflator.
- Specific production issues:
  - Owner-occupied dwellings: output valued as estimated rentals tenants would pay; benchmark data are outdated, leading to biased estimates.
  - Work-in-progress: Growing crops, standing timber, stocks of fish, and livestock reared for food are not treated as work-in-progress; total output allocated to the period of harvest or slaughter.
  - Inventory valuation adjustment: Inventories are not adjusted for holding gains/losses; estimates deflated using the general wholesale price index.
  - Consumption of fixed capital: Perpetual inventory method is not used; experimental estimates compiled and plan to introduce the method with revised GDP on new base year.
  - Cash vs. accrual: Government transaction data received on a cash basis are not adjusted to accruals.
- Volume measures and deflation practices:
  - Base year for constant price estimates is 1980/81.
  - International best practice recommends chain linking; otherwise base year should be updated at least every five years.
  - Deflators are compiled at a highly aggregated level; many indicators do not reflect underlying price changes of aggregates.
  - Wholesale price index is widely used but may include transportation and trade margins not part of output value.
  - Double deflation techniques are not used; system depends heavily on reflation of volume estimates to derive current price estimates.
  - Volume measures of taxes and subsidies on products are estimated by deflating current estimates using the wholesale price index, contrary to 1993 SNA recommendations.

### Expenditure approach procedures
- Household final consumption expenditure is derived as a residual; FBS cites difficulty obtaining data for a commodity flow approach.
- Government final consumption expenditure is estimated at the one digit level of the COFOG. Constant price estimates use a wage index for compensation of employees and wholesale price index for other goods and services.
- Gross fixed capital formation (GFCF) is estimated by type of economic activity for major groups and by type of asset; sources are similar to value added estimates and affected by same data deficiencies.

### Other statistical procedures and treatment of informal economy
- Techniques for data adjustments and transformations are weak.
- Procedures to obtain exhaustive measures of GDP are not employed despite high informal activity, unrecorded cross-border trade, and under-coverage of periodic surveys.
- FBS is receiving assistance from the United Nations Economic Commission for Asia and the Pacific to derive preliminary estimates of informal activity.

### Assessment and validation of source, intermediate data, and outputs
- Source data assessment:
  - Sampling errors, response errors, and nonsampling errors are routinely assessed, but the information is not made available to users; it is used internally to inform estimation.
  - Data are checked for errors and inconsistencies using range checks and benchmarks.
  - A three-year project (July 1995 to June 1998) was conducted to assess and validate FBS statistics with objectives to improve validity through cross-checking, analyze data from various sources, and conduct surveys and studies in weak areas.
- Validation of intermediate data and outputs:
  - Production of primary goods is assessed against merchandise trade statistics and other secondary sources, including HIES.
  - Ad hoc assessments are conducted when discrepancies in intermediate data are identified.
  - The statistical discrepancy between GDP by activity and GDP by expenditure is not shown separately because private final consumption expenditure is derived as a residual; FBS cannot conduct studies of statistical discrepancy.
  - Supply and use tables have been compiled but are not recent and the framework is not used to investigate discrepancies.
  - FBS does not compare its estimates with occasional independent GDP estimates from research institutions.

*Source: _cr04398 - 2.3       Classification/sectorization*

### 3.5 Revision studies

### 3.5 Revision studies

### Studies and analyses of revisions
- As part of the comprehensive plan to improve the estimates, the FBS commissioned 23 studies to review data sources and statistical techniques used to compile estimates of output, value added, and gross fixed capital formation for the broad range of economic activities covered by the compilation process.
- The studies introduced experimental estimates based on new data sources and techniques.
- The first study was completed in May 2001 and the last one in May 2002.
- The results of these studies have been used as part of the exercise to rebase the estimates to 1999/2000.
- Revision studies have been made available to national and international research agencies for review and comment, and are available to the public on request.

### Serviceability — relevance and user engagement
- The NAC is active in reviewing national accounts statistics and encompasses a broad range of users from both the public and private sectors.
- When compiling statistics, the FBS considers comments received during the press conference at which estimates are presented.
- The FBS reports it receives very little feedback from users not represented on the NAC.
- FBS staff attend international meetings and seminars where issues relating to the statistics are discussed.

### Timeliness and periodicity
- The first GDP releases are published about two weeks before the end of the fiscal year and are based on nine to ten months’ observed data and two to three months’ projections.
- The first annual GDP data based on 12 month’s observations are disseminated with a lag of 11 months (vis-à-vis the recommended 6-9 months).
- The FBS notes this situation would be remedied with the publication of quarterly national accounts as presently in preparation.
- Estimates of GDP are compiled on an annual basis, which is in line with the recommendations of the GDDS.

### Consistency and reconciliation
- The statistical discrepancy between estimates by type of activity and category of expenditure is not shown independently, but is included in private final consumption expenditure as a residual; therefore consistency between the two estimates cannot be investigated.
- Consistent estimates of GDP by type of activity are available from 1950/51 and by category of expenditure from 1964/65.
- The FBS adjusts back series, wherever possible, to take account of changes in methodology to eliminate discontinuities in the data.
- The national accounts statistics are reconcilable with the balance of payments statistics, the foreign trade statistics, and the government finance statistics.

### Revision policy and practice
- The revision process follows a regular, well-established schedule because the FBS is obligated to provide data for inclusion in the economic survey compiled by the MOF.
- Provisional estimates compiled one month before the end of the reference period are updated six months after the end of the period, but the updated estimates are submitted only to government agencies.
- A comprehensive revision of the provisional estimates is undertaken and these data are released eleven months after the end of the period.
- Preliminary data are clearly identified as such in press releases, on the website, and in hardcopy publications produced by the FBS.
- The FBS claims there is no public interest in the updated estimates released to government agencies, though some users have complained about the lack of regularly revised data.
- The revision cycle is not known to users.

### Accessibility of data and metadata
- Statistics are presented in the national accounts publication, on the website, and in the yearbook with charts, tables, and explanatory notes; the national accounts publication includes summary tables with time series and detailed tables.
- Data are first presented via a press release and are subsequently posted on the FBS website; detailed data are presented in the annual national accounts publication National Accounts of Pakistan and other hardcopy publications.
- The FBS’s advance release calendar does not include release dates for the national accounts; users are aware of the timeframe by tradition and practice, but the precise date may vary.
- Data are presented to the National Accounts Committee before finalization and wider dissemination to the public.
- Nonpublished (but nonconfidential) data are available to users for a fee; micro-level data may be supplied on tapes/diskettes after publication subject to assurances on third-party distribution and acknowledgement requirements.
- Documentation on methodology, data sources, and statistical techniques is provided in the annual national accounts publication and on the FBS website; metadata are also available on the DSBB.
- The metadata do not provide sufficient information on deviations from international guidelines or the weaknesses in source data; the national accounts publication does not detail possible biases in source data, the degree of coverage, or the level of non-response.
- The 23 case studies describe current practices and contain a discussion of the weaknesses.
- Details about concepts, definitions, sampling procedures, and other attributes of surveys and censuses are provided in the reports of those surveys; references are presented in the national accounts publication.

### Assistance to users and dissemination practices
- Publications and the website list names, email addresses, and telephone numbers of FBS staff, but do not specify the subject field of the contact person.
- The FBS compiles a catalog that lists the price and frequency of available publications; the hardcopy catalog is published annually. The website catalog does not list prices.
- The FBS operates a bookstore at its headquarters in Islamabad.

### Key assessment findings (excerpted from Table 1)
- Accounts are based on 1968 SNA.
- COICOP not used; 1993 SNA not used to classify units.
- Market output recorded at factor cost; sales taxes excluded from intermediate consumption.
- Enterprise data weak and outdated; computing resources obsolete.
- Timeliness not in line with GDDS; FBS expected to disseminate quarterly GDP data by activity and category of expenditure in early 2004.
- Contact information not specific to subject areas.

### Recommendations relevant to revisions, accuracy, and serviceability
- Formulate plans for the next rebasing exercise; estimates should be rebased with a minimum frequency of five years (following release of the new series at 1999/2000, the estimates should then be rebased to 2004/05 at the latest).
- Undertake regular (preferably annual) enterprise surveys taking account of all formal activities not covered by administrative statistics to improve source statistics for business services.
- Derive independent estimates of private final consumption expenditure using HIES data to enhance consistency checks and improve overall estimates.
- Disseminate to the public the statistics that are revised six months after the end of the reference period.

*Source: _cr04398 - 3.5 Revision studies*

### 0.2       Resources

### 0.2       Resources

### Staff, financial, and computing resources
- The CPI, SPI, and WPI are compiled by the price statistics section, located in Karachi and headed by a director.
- Centrally, 22 staff are employed, of which 12 are statistical officers and six statistical assistants.
- All statistical officers are postgraduates in statistics, mathematics, economics, commerce, or sociology.
- New staff receive on-the-job training; sitting staff have limited opportunities for training limited to computer use or developments in price statistics per the revised international CPI and PPI manuals.
- Staff in the 35 FBS regional and field offices spend a large portion of their time on price collection; local staff receive training via the FBS Manual of Instructions for Price Collection.
- Supervisors received 15 days training in the FBS Lahore training wing and at Karachi on the occasion of the recent change of the base years.
- Staff turnover is very low; a core contingent of trained staff is therefore generally assured.
- Motivation issues among staff include: a promotion policy that gives little weight to individual performance, low mobility possibilities within the civil service except for the highest grades, insufficient allowances for travel expenses, nonindexation of housing allowances, and repeated refusal of annual leave requests.
- Transition plans: decision to transform the Statistics Division (which includes FBS, Agricultural Census Organization, Population Census Organization) into a more autonomous agency may open improvement possibilities.
- Computer resources at Karachi headquarters are hardly sufficient; most calculations are done on a mainframe that is rapidly becoming obsolete.
  - There is no database for prices: storage is done on diskettes.
  - The number of personal computers is low and insufficient for the envisaged producer price index that should run on PCs.
  - Regional and field offices: only a few have a PC and can transmit price information by e-mail; most offices transmit written overviews by fax; some smaller offices still depend on express mail delivery.
- Financial resources are overall sufficient for the price index program, though actual costs are usually higher than the budget permits.
  - Exception: insufficient allowance for travel expenses, which especially affects price collectors in regional and field offices.

### Measures to ensure efficient use of resources
- Budget control is strict in the Pakistan civil service; efficient use of resources is a requirement.
- Tedious clearance procedures and arrears in payment of budgeted allowances sometimes reduce the Statistics Division’s ability to respond dynamically to efficiency improvements.
  - Example: difficulty obtaining Ministry permission to equip price enumerators with motorcycles despite limited budgetary consequences.
- Consistency in concepts and methodologies across FBS units avoids unnecessary recoding and adaptation.
- Data procedures are managed to minimize processing errors; coordination meetings are regularly held to monitor work processes.
- Overall resource usage of price statistics is not known (footnote: from time to time an estimate is made of the expenditure related to the price collection activity of regional and field offices).
- Efficient work processes are supported by an extensive manual of operating instructions.

### Quality awareness

- Management is sensitive to all dimensions of data quality; organizational infrastructure supports data quality as reflected in the mission statement emphasizing accurate, reliable, timely, comprehensive and user-friendly information, research, up-to-date technical knowledge, and adoption of improved methodologies and technologies.
- Special processes focusing on quality such as Total Quality Management, ISO 9000, or external audits are not used.
- Reviews are undertaken to identify problems in collection, processing, and dissemination.
  - Guidance comes from the NSC and its subcommittees.
  - For price statistics, an experts group chaired by the chief economist of the Planning and Development Division provides guidance for important changes (e.g., change of base year).
  - A standing panel on price statistics convenes as requested by FBS; comprises about 20 main users and stakeholders and is chaired by the Secretary of the Statistics Division.
  - Further feedback: workshop for users of social statistics and occasional questionnaires added to the Pakistan Statistical Yearbook or specially sent to users.
- Trade-offs among quality dimensions are recognized; in practice timeliness prevails over accuracy/reliability—this choice is known to advisory bodies on price statistics.
- Planning for quality improvements evidenced by:
  - Enhancements in data dissemination (Statistics Division’s website, revamped Monthly Bulletin of Statistics, new Pakistan Statistical Pocket Book).
  - Plans to develop professional skills via extension of the training wing.
  - Continuing improvements on the CPI and development of a producer price index.
- Price statistics section responds to user requirements: recent changes included widening geographical scope of CPI, additional commodity breakdowns, modified treatment of seasonal items, changes in treatment of rents, and renaming “medicines” to “Medicare.”

### Integrity — Professionalism and transparency
- The General Statistics Act does not contain a provision regarding professional independence of the data producing agency.
- The predominant function of the Secretary of the Statistics Division in FBS decision making could allow political interference; however, there is no recent evidence of such interference.
- A 1996 task force reported credibility concerns tied to lack of transparency and non-publication of methodologies; since 1996 FBS has made progress, notably via its website.
- Price bulletins are signed off by the FBS director general and are not reviewed by any other party before publication.
- Professional competency is key in recruitment but less so for promotions; staff are appointed permanently after meeting civil service requirements.
  - The director general must be appointed from staff ranks and can remain until retirement.
  - Professionalism promoted through contributions to methodological descriptions, organization of conferences/workshops, and permission to attend local economic association meetings.
- FBS sends letters of clarification and requests rectification upon discovering erroneous interpretation or misuse of data; it verifies for internal errors first.
  - Occasionally organizes press briefings and has in the past organized a workshop for users of statistics.
  - The Coordination section maintains a clipping service.
- The Statistics Division’s website contains the mission statement and the text of the Statistics Act, 1975.
- FBS publications generally include contact information; data dissemination policy outlined in the Brochure of the Federal Bureau of Statistics 2002-2003 on the FBS website.
- There is no government access to price statistics prior to the date of release.
- Data released are clearly identified by the Statistics Division’s and/or the FBS’s name; FBS requests attribution when its statistics are reproduced.
- Advance notice of major methodological changes is given to newspapers and appears in the Statistics Division’s newsletter; the website is also recently used for announcements.
- Ethical rules:
  - The General Statistical Act strictly prohibits divulging individual data.
  - Civil Service Rules of 1973 include more extensive conduct rules and a copy is available in every FBS office, but these are not widely known by staff.
  - No further written ethics rules at the Statistics Division level; staff are not regularly reminded of rules.
  - Proper conduct is expected and periodically monitored individually.
  - The Manual of Instructions for Price Collectors emphasizes politeness, explaining the purpose of price collection, and these rules are followed up in training.

### Methodological soundness — Concepts, scope, classification, recording
- CPI measures changes in the cost for urban households of buying a fixed basket representing all types of final consumption expenditure.
- Four income brackets are distinguished; previously price indices for different categories of employees were computed.
- Overall CPI concepts and definitions follow International Labor Organization guidelines; transaction characteristics are taken into account when defining prices collected each month.
- Scope:
  - CPI scope limited to households living in urban centers; excluded are households in areas under military control.
  - Families of all sizes, economic activities, and income levels are covered.
  - Scope of transactions matches household consumption expenditure per 1993 SNA: monetary purchases, goods/services for final consumption, goods produced on own account, receipts of assistance in kind, and services of dwellings occupied by owners.
  - Illegal goods and services (such as wine) are not included.
- Classification:
  - Neither COICOP nor CPC is used.
  - An eight-digit national classification is used: first two digits indicate national consumption function; second two digits the type of commodity (manufactured ones classified per Pakistan Standard Industrial Classification); third pair the item; last two digits the urban center.
  - Deviation from international classifications is kept under review with consideration of introducing COICOP and/or CPC.
- Basis for recording:
  - Market prices valued at local market prices including trade and transportation margins and applicable taxes less subsidies.
  - Cash purchases with immediate delivery are recorded; sale prices are included except when they reflect an abnormal situation.
  - If an item is controlled via ration stores and sold in the open market, the open market price is recorded.
- Recording schedule (all prices used for a month's CPI are collected in that month):
  - Food and beverages: collected during the 11th–14th of the month.
  - Apparel, textile and footwear, fuel and lighting: collected during the 1st–3rd of the month.
  - Household furniture and equipment: collected during the 4th–6th of the month.
  - Recreation, entertainment and education, cleaning, laundry and personal appearance and Medicare: collected during the 7th–10th of the month.
- Grossing/netting:
  - CPI weights treat purchase of existing goods on a gross basis instead of net of sales in secondary markets, in variance with 1993 SNA recommendations.

### Accuracy and reliability — Source data and collection
- CPI weights derived from the Family Budget Survey conducted in urban areas during 2000-2001.
- The Family Budget Survey is not annual but compiled with a frequency appropriate to changes in consumer behavior; lack of annual frequency prevents compiling a chain index.
- Institutional coverage includes all four provinces (except parts under military restrictions).
- Sample design for Family Budget Survey:
  - Area frame divided into enumeration blocks of about 200-250 households.
  - Total of 22,800 urban enumeration blocks distinguished.
  - 2,053 primary sample units in 52 cities or towns were drawn by probability proportional to size (PPS), considered sufficient to produce estimates at national and provincial level at 95 percent margin of confidence.
  - From these primary sample units, 20 households selected by random systematic sampling.
  - Total sample size was 41,060 households; about 10,000 households were surveyed each quarter of fiscal year 2000-01.
- Replacement policy: if selected households migrated from the primary sample unit, a replacement was sought. Nonresponse was very low (less than 1 percent) and did not lead to replacement to keep sampling design intact.
- Data collected are sufficiently detailed and include rental value of owner-occupied dwellings.
- Price collection:
  - Most prices collected through a monthly price collection survey; coverage is comprehensive including all types of consumer acquisitions.
  - Stratified random selection of urban centers by size category plus an urban center of less than 50,000 inhabitants in each province generated 52 centers; final selection for price collection included only 35 cities based on marketing of items, similar price trends in small centers, and cost considerations.
  - Other selections are purposive: 71 markets selected on volume of sales and income groups frequented; 374 items selected on general availability in urban centers; outlets selected on volume of sales.
  - From each market and item, four quotations are obtained; usually the simple average of those four is taken as representative price for that market. City average price for each item obtained by averaging selected markets in that city.
  - Price data are sufficiently detailed by product with sufficient product specifications to ensure same variety is priced each period.
  - No ad hoc surveys are conducted to supplement the regular survey program (e.g., to identify new products).
  - Some prices collected directly for items under government control or with a single provider (electricity, water, telephone and train fares).
- Price developments monitored through contacts with market committees, newspapers, etc.
- Source data reasonably approximate definitions, scope, classifications, valuation, and time of recording required.
- Source data are timely for monthly CPI calculation.

*Source: _cr04398 - 0.2       Resources*

### 3.2       Statistical       techniques

### 3.2       Statistical       techniques

### 3.2.1 Data compilation employs sound statistical techniques

- The level of detail for commodities is sufficient for analysis; at least the ten largest groups in the classification are used.
- Weights in the overall CPI reflect the aggregate value of the commodity group acquisitions in the base year in the 52 cities that were selected through stratified random sampling; weights can be seen as representing the total urban population, although it might have been preferable to take the weight totals directly from the Family Budget Survey. No corrections are made for underreporting of certain types of expenditures.
- Housing measurement:
  - For both rented and owner-occupied housing, price changes are computed on the basis of eight indices representing construction costs.
  - To take out undue fluctuations, these data have been smoothed by taking the geometric mean for the past 24 months.
  - Apart from the smoothing, this procedure does not reflect internationally accepted standards, as construction costs are a too indirect measure of actual rentals (or, in the case of owner-occupiers, the estimated rentals that tenants would pay for similar accommodation).
  - The root problems: most households live in dwellings they own (rental market is thin) and contractual arrangements fix prices for long periods, resulting in unacceptable results when trying to measure price change through surveys.
  - In the Family Budget Survey, the imputed rentals for owner-occupied dwellings are based on self-assessments by the owners.
  - The weight of housing in the total CPI is 23.4 percent.
  - Proposed common solutions: start recording changes in the purchasing price of dwellings or, if no better alternative, delete the housing item completely from the scope of the CPI.
- Calculation of elementary indices:
  - An unbiased formula is used, namely the ratio of average prices.
  - Average prices are mostly determined as the arithmetic average of the observed quotations. Where the spread in prices is quite large, a geometric average is taken.
  - To compute the national CPI, the unweighted average of quotations in all selected markets is taken—acceptable if the CPI represents the selected cities rather than the full urban population.
- Aggregation and base/reference periods:
  - A fixed-based Laspeyres index formula is used to aggregate elementary indices to higher levels.
  - The weight reference period and the price reference period are the same: the full year 2000-2001.
  - Use of a full year as price reference period is acceptable, but most countries prefer a particular month (for example, June 2001) to avoid a certain jump between the end of the base year and the start of the next year (in Pakistan’s case between June and July 2001).
  - The weight reference year is very recent. When new weights are introduced, the FBS constructs backward series for main categories and the overall CPI by splicing over the annual data.
- Treatment of missing items and substitutions:
  - Temporarily missing prices are handled by carrying the last reported price forward for a short period; if the item remains unavailable in the selected outlet, another outlet is chosen.
  - When a product disappears from all markets, normally a close substitute is selected. In case of substitution, a base year price for the new item is imputed by applying the price change since the base year for the old item to the current price of the new item.
  - If called for, appropriate adjustments are made for quality differences. The FBS does not add new products to the sample between changes of the base year.
- Seasonal products: monthly weighting schemes have been developed that leave the weight for the main commodity group unaffected.

### 3.2.2 Other statistical procedures employ sound statistical techniques

- The 52 selected urban centers are representative of about 68 percent of the total population.
- The implicit deflators in the national accounts are not an alternative for the part not covered by the CPI because national accounts data for household consumption are calculated as a residual.

### Assessment and validation practices (related sections 3.3–3.5 and 3.4)

- Source-data assessment:
  - Sampling errors for the Family Budget Survey were calculated; these data are analyzed. A post-enumeration survey was carried out.
  - Influence of outliers in price quotations is reportedly limited; high-value transactions and unlikely quotations are verified.
  - Regional and field-office supervisors must verify each month 10 percent of recorded price quotations; enumerators make notes in a remarks column (e.g., if there is a variation of 5 percent or more in prices compared to the previous month, specific reasons should be given).
  - Supervisors’ findings are laid down in written reports. Accuracy is also verified in the Karachi office of the FBS. Computer-assisted verification focuses on temporal and geographical consistency; standard deviations are computed. Consistency with other sources (newspapers) is reviewed. Inspection visits are undertaken where persistent problems are noted; no external audits take place.
  - Source data are especially scrutinized in the context of a change in the base year.
- Validation of intermediate data and outputs:
  - The CPI is compared with the SPI and the WPI and validated against other sources such as newspapers, the Bureau of Supply and Prices, and Vital Information Service (prices in Karachi and Lahore).
  - Once prices are verified, they are sent to the computer service for index calculation; indices are checked by a number of agents before submission to the Director-General for final decision.
- Revisions:
  - On change of the base year, the Experts Group analyzes revisions; the relevant report is printed and available for interested users.
  - Occasionally, if errors are found, the CPI is revised—historically this has happened to the data for three months back. This informs statistical processes and serves as a warning to be more vigilant in certain areas.
- Recommendation captured elsewhere: consider using geometric averages instead of arithmetic averages for all elementary indices.

### Serviceability highlights (sections 4.1–4.4)

- Relevance and user engagement:
  - Monitoring is done largely via the Experts Group and the Panel for Price Statistics, which include users and other stakeholders; feedback also comes from workshops, seminars, and user questionnaires.
  - CPI is a standard item on the agenda of the monthly meetings of the Economic Coordination Committee of the Cabinet (secretary of the Statistics Division and the FBS Director-General participate).
  - FBS participation in international meetings on price indices is limited; normally the Director-General attends.
- Timeliness and periodicity:
  - The timeliness of the CPI is in accordance with the GDDS as well as the SDDS.
  - The periodicity of the CPI is in accordance with the GDDS as well as the SDDS.
- Consistency:
  - Estimates for all classification typologies (commodities, geographical areas, and income groups) are consistent: the aggregate is invariant to typology of aggregation.
  - The CPI is published in the Pakistan Statistical Yearbook over a consistent period of ten years.
  - On request, data are provided in a 59 item breakdown.
  - Consistent CPI series showing price developments over half a century were constructed on the occasion of the FBS’s 50-year jubilee.
  - The CPI is consistent with the national accounts in that index data are used to deflate national accounts household final consumption expenditure.
- Revision policy and practice:
  - Thus far, the CPI has been revised with a periodicity of every ten years; a reduction to every five years is being considered.
  - The most extensive report on the change of the base year (by the Experts Group) is available to the public.
  - Monthly CPI data are not normally revised; consequently, there are no preliminary data. If a mistake is rectified, all previously published detail is again disseminated with a note explaining the reason for change.
  - Studies and analyses of revisions are made public; the Experts Group report on the change of the base year is made public after approval by the Minister of Finance.

### Accessibility and user support (sections 5.1–5.3)

- Data presentation and publications:
  - The Monthly Review of Price Indices (first release) has a clear layout of text and tables; CPI, SPI, and WPI are presented side-by-side. Tables give changes compared with one month back and one year back and include index weights. A short commentary is included; some publications include a few graphs.
  - The Family Budget Survey 2000-2001 has not been published; CPI metadata contain few details on this survey. A five-page description of the sample design can be obtained from the FBS Sampling section on request.
  - Data are first released in the five-page Monthly Review of Price Indices, followed by several publications (including the Monthly Bulletin of Statistics, which is free and available on the Internet) that present extensive detail.
- Release schedule and equality of access:
  - The CPI is released on the 11 th of the month following the reference month according to a preannounced schedule.
  - Hardcopies are released to selected users within government and to the press shortly before publication on the FBS website.
  - Nonpublished (but nonconfidential) subaggregates are made available upon request; this opportunity is not widely publicized but main users have been informed by letter.
- Metadata and assistance:
  - Metadata on FBS price indices are published on the Statistical Division’s website, in the annual Brochure of FBS, the Pakistan Statistical Yearbook, the July and December issues of the Monthly Statistical Bulletin, and on the IMF Data Standards Bulletin Board. A detailed methodological note is available on request.
  - Slightly different versions of metadata exist; the annual Brochure of FBS version is best suited for a general public.
  - Most FBS publications contain general contact information; some include names and telephone numbers of directors. The Monthly Bulletin of Statistics includes no names or telephone numbers of price-statistics specialists, but prompt and knowledgeable support is available.
  - A catalog of publications and some documents is available (including on the Internet); a catalog of other documents and services does not exist. Assistance with placing orders is available from the National Book Foundation. The catalog includes many very old publications and is not always up-to-date.

### 0. Prerequisites of quality

### 0. Prerequisites of quality

### Integrity
- See section on the consumer price index.

### Methodological soundness
- Concepts and definitions
  - The wholesale price index (WPI) is designed to measure the directional movement of prices in the domestic primary and wholesale markets.
  - The overall structure of the WPI concepts and definitions are in accordance with internationally accepted standards.

- Scope
  - The scope of the WPI covers the full country and five main commodity groups:
    - food;
    - raw materials;
    - fuel, lighting & lubricants;
    - manufactures, and
    - building materials.
  - Owing to practical difficulties, the weights for manufacturing are based on enterprises with 10 or more workers.

- Classification/sectorization
  - The classification used is the Pakistan Standard Industrial Classification, which is derived from the International Standard Industrial Classification of All Economic Activities (ISIC), Rev.2.
  - It has been decided to change to ISIC Rev. 3 in the near future.

- Basis for recording
  - Market prices used are those realized by primary sellers or going at organized wholesale markets; they normally include applicable transportation costs and (local) taxes, except for the national sales tax, which is considered a tax imposed on final customers.
  - For agricultural products, an adjustment for transportation may be made to bring the observed wholesale prices closer to the price received by the farmer in the primary market.
  - The prices recorded should reflect cash transactions with immediate delivery. In several cases, list prices are accepted.
  - Recording is done on an accrual basis: all prices used for the calculation of the WPI for a particular month are collected during the 13th through the 15th of that same month.
  - Grossing/netting procedures: Not applicable.

### Accuracy and Reliability
- Source data
  - WPI weights reflect the value of marketable surplus in the base year, marketable surplus being defined as the sum of output value (less final consumption of own-account production in the case of agricultural products) and imports, less exports.
  - Principal sources for the weights are statistics on agriculture, the Census of Manufacturing Industries (CMI), and foreign trade statistics.
    - The CMI is limited to establishments with more than 10 workers.
    - Foreign trade statistics do not reflect considerable smuggling that is generally accepted to take place in certain types of commodities.
  - The basket includes 106 commodities and 425 items.
  - Monthly price collection survey:
    - Based mainly on observations during personal visits of FBS enumerators to primary wholesale markets in 18 cities.
    - For each item, four quotes are collected, but only in the principal market for the item in question.
    - Prices for certain items are directly reported by producers or importers (petroleum products, cars, etc.), or obtained from trade associations (cotton, timber, etc.).
    - Markets and items selected on trade volumes; no random sampling techniques have been used.
    - If prices are collected from producers rather than markets, only a single quote is available; usually list prices are accepted. Several producer-provided prices have shown no changes for long periods (up to eight years), raising concerns that sometimes not actual prices have entered the index.
  - No ad hoc surveys are conducted to supplement the regular survey program.
  - Coverage: Reportedly, the proportion of domestically traded goods falling in the definition of the WPI that is not covered by the index is about 20 percent.

- Statistical techniques
  - Elementary indices: simple arithmetic averages of prices observed in all markets where items are traded are compared with the average price in the base year; a geometric mean is taken if prices fluctuate importantly (e.g., cotton).
  - Aggregation: fixed-based Laspeyres index used. The weight reference period and the price reference period are the same: the full year 2000-2001.
    - Note: Use of a full year as price reference is acceptable, but most countries prefer a particular month (say June 2001) as price reference to avoid a certain jump between the end of the base year and the start of the next year (in Pakistan between June and July 2001).
  - Weight reference year is very recent; when new weights are introduced, the FBS constructs backward series for main categories and the overall WPI by splicing over the annual data.
  - Temporarily missing prices: last reported price carried forward for a short period; if unavailable for more than two months, a close substitute is selected. In case of substitution, a base year price for the new item is imputed by applying the price change since the base year for the old item to the current price of the new item. Appropriate adjustments are made for quality differences if called for.
  - FBS does not add new products into the sample between changes of the base year.
  - Seasonal products: monthly weighting schemes developed that leave the weight for the total product group unaffected.

- Assessment and validation of source data and outputs
  - Outliers: high-value transactions and unlikely quotations are verified and deleted if no satisfactory reason is found.
  - Supervisors at regional and field offices verify each month 15-20 percent of recorded price quotations; findings are laid down in written reports.
  - Computer-assisted verification in Karachi office focuses on temporal and geographical consistency; standard deviations are computed; consistency with other sources (newspaper reports) is reviewed.
  - Inspection visits undertaken to regional or field offices when persistent problems are noted; no external audits take place.
  - Source data are especially scrutinized when changing the base year.
  - WPI is compared with the SPI and the CPI; validation also done with other sources such as Bureau of Supply and Prices, market committees, and newspapers.
  - Indices are checked by a number of agents before submission to the Director-General for final decision.
  - Revisions: On occasion of the change of the base year, the Experts Group analyzes the revisions; the relative report is printed and available for interested users. Occasionally the WPI is revised (in the past this has happened to the data for three months back).

- Revision studies
  - The Experts Group analyzes revisions on change of base year and produces a report. Errors have led to revisions up to three months back in the past; this serves as a warning to be more vigilant.

### Serviceability
- Relevance
  - Monitoring of relevance and practical utility takes place largely via the Experts Group and the Panel for Price Statistics, comprising users and public and private stakeholders.
  - Feedback also received via workshops, seminars, user questionnaires.
  - WPI is a standard item on the agenda of monthly meetings of the Economic Coordination Committee of the Cabinet.
  - FBS participates to a limited extent in international meetings on price indices; normally the Director-General attends.

- Timeliness and periodicity
  - Timeliness of the WPI is in accordance with the GDDS as well as the SDDS.
  - Periodicity of the WPI is in accordance with the GDDS as well as the SDDS.

- Consistency
  - Estimates produced are internally consistent.
  - WPI is published in the Pakistan Statistical Yearbook over a consistent period of ten years.
  - WPI is consistent with the national accounts in the sense that index data are used to deflate various items in the accounts.

- Revision policy and practice
  - Thus far, the WPI has been revised with a periodicity of every ten years; a reduction to every five years is being considered.
  - The most extensive report on the change of the base year (the one by the Experts Group) is available to the public (after approval by the Minister of Finance).
  - Preliminary data: monthly data of the WPI are seldom revised; as a consequence, there are no preliminary data. If a mistake is rectified, all details as previously published are again disseminated, accompanied by a note explaining the reason for the change.
  - Studies and analyses of revisions are made public; users are informed of causes of revisions.

### Accessibility
- Data accessibility
  - The first release regarding prices, the Monthly Review of Price Indices, has a clear layout of text and tables. CPI, SPI, and WPI are presented side-by-side. Tables give changes for the most recent period in comparison with one month back and one year back; a short commentary is included. The tables provide information on the index weights. Some publications include a few graphs.
  - Data are first released in the five-page Monthly Review of Price Indices, followed by several publications, some of which (including the Monthly Bulletin of Statistics, which is free and available on the Internet) provide extensive detail.
  - The WPI is released on the 11th of the month following the reference month according to a preannounced schedule.
  - Hardcopies of the data are released to selected users within the government as well as to the press shortly before they are published on the FBS website.
  - Nonpublished (but nonconfidential) subaggregates are made available upon request; this opportunity is not widely publicized, but main users have been made aware of it in a letter.

- Metadata accessibility
  - Metadata on the FBS price indices are published on the Statistical Division’s website, in the annual Brochure of FBS, the Pakistan Statistical Yearbook, the July and December issues of the Monthly Statistical Bulletin, and on the IMF Data Dissemination Bulletin Board.
  - A detailed methodological note is available on request.
  - Slightly different versions of the metadata exist; the one in the annual Brochure of FBS is best suited for a general public.

- Assistance to users
  - Most publications of the FBS contain general contact information; some include names and telephone numbers of directors. The Monthly Bulletin of Statistics includes no names and telephone numbers of specialists in price statistics. Prompt and knowledgeable support is available to users of the WPI.
  - A catalog of publications and some other documents, such as training notes, is available, including on the Internet. A catalog of other documents and services does not exist. Assistance with placing orders is available from the National Book Foundation. The catalog includes many very old publications and is not in all cases kept up-to-date.

### Key findings, issues, and plans for improvement (from Table 3 summary)
- Prerequisites of quality
  - Legal and institutional environment: assessment = O (Practice Observed). Plan: Legislation to give more autonomy to the FBS (short tem).
  - Resources: assessment = LO (Practice Largely Observed). Comment: There is room for improvement of staff motivation and computer resources.
  - Quality awareness: assessment = X (marked in table).

- Integrity
  - Professionalism: assessment = X.
  - Transparency: assessment = X.
  - Ethical standards: assessment = X.

- Methodological soundness
  - Concepts and definitions: assessment = X. Plan: Publish Producer Price Index (short term).
  - Scope: assessment = X.
  - Classification/Sectorization: assessment = X.
  - Basis for recording: assessment = X.

- Accuracy and reliability
  - Source data: assessment = LO. Comment: Lack of robustness in the sources for the weights.
  - Statistical techniques: assessment = X.
  - Assessment and validation of source data: assessment = X.
  - Assessment and validation of intermediate data and statistical outputs: assessment = X.

*Source: _cr04398 - 0. Prerequisites of quality*

### 3.5             Revision             studies

### 3.5 Revision studies

### Serviceability — data dissemination and revisions
- Some users receive data shortly in advance of public dissemination.
- Contact persons are not always publicized; catalog of services is not available.
- Recommendations regarding the Wholesale Price Index are located on the pages directly following Table 2.

### Prerequisites of quality — institutional and legal arrangements
- No specific legal responsibility has been granted to any agency or government unit to compile government finance statistics (GFS) for the consolidated general government sector.
- The Ministry of Finance (MOF) has accepted responsibility for compiling GFS for the budgetary sector under the general provisions of the federal government Rules of Business 1973 for the Finance Division.
- MOF compiles and disseminates quarterly and annual GFS for the budget sectors for all levels of government in the Economic Survey and on its website.
- Quarterly data on central government gross debt are published on the SBP website.
- MOF has not taken responsibility for compiling GFS for non-budget agencies.
- There is no dedicated division in the MOF responsible for the compilation of GFS; compilation is carried out by staff in addition to their regular responsibilities.
- Provision of revenue and expenditure data by AGPR/AGs, and financing data by the SBP, is based on informal arrangements; no legal mandate exists to obtain these data.
- There is no legal mandate and/or measures to encourage reporting of data for extra-budgetary units to the MOF, preventing compilation of GFS covering all general government units.

### Data sharing, coordination, and compilation practices
- Responsibility for preparation of government accounts:
  - Accountant General for Pakistan Revenue (AGPR) for the central (federal) government.
  - Provincial accountant generals (AGs) for provincial and local government levels.
- AGPR compiles GFS revenue and expenditure statistics for the central government budget and provides these data to MOF on a monthly basis.
- Provincial government data are compiled by provincial AGs and provided to MOF and AGPR on a monthly basis; provincial GFS include local government data down to the district level.
- GFS domestic financing data are compiled by the State Bank of Pakistan (SBP) and provided to MOF monthly, but with a lag of two months.
- External financing data are compiled by the External Affairs Division of MOF.
- GFS debt data are compiled by the Debt Management Unit (domestic debt) and External Affairs Division (external debt) of MOF.
- The Economic Advisor wing of MOF coordinates provision of government finance data to the Federal Statistical Bureau for national accounts compilation.
- Reconciliation of accounting data is overseen by the Fiscal Monitoring Committee (FMC), which comprises representatives from MOF, AGPR, SBP, Central Board of Revenue (CBR), External Affairs Division of the MOF (EAD), and the National Savings Schemes (NSS).

### Resources and capacity
- No staff in MOF, AGPR, or SBP are assigned specifically to GFS compilation; compilation is done in addition to regular tasks.
- Sufficient staff and computing resources are available in MOF to compile GFS on the current limited basis, but:
  - A need exists for more extensive statistical training of staff, who presently rely on on-the-job training focusing mainly on processing procedures.
  - AGPR and provincial AGs need better computer resources and systems and extensive GFS statistical training.
- Existing resources are sufficient for current compilation that covers summary operations of the budgetary government, but insufficient for compilation of consolidated general government data covering the full range of GFS classifications.
- A dedicated GFS group is required to develop an output and user focus and to improve serviceability and accessibility of GFS.
- There is no separate cost information on the compilation of GFS; resource allocation is part of overall management processes in MOF, AGPR/AGs, and SBP.

### Quality awareness and monitoring
- Focus is on application of accounting rules and established procedures; compilers lack a user perspective, so quality focus is on procedures rather than products.
- Processes exist to monitor quality of collection and processing: monthly accounting data production by AGPR and provincial AGs, followed by inter-agency reconciliation overseen by the FMC.
- Fiscal data are initially compiled using preliminary and un-audited data, then recompiled against audited data.
- The discrepancy between the deficit/surplus derived from revenue and expenditure data and that derived from financing data is monitored and reconciled as far as possible.
- Planned improvements to government accounting standards have taken account of international accounting standards (IFAC) and are generally consistent with the GFSM 2001 classification structure.
- No systematic process is in place to obtain feedback from users of GFS; contact details are provided on Pakistan’s GDDS website but not in the Economic Survey or on the MOF website.
- No specific processes are in place to deal with quality considerations or statistical planning; trade-offs between timeliness and accuracy are implicitly managed by releasing preliminary statistics followed by statistics based on audited data.

### Integrity and transparency
- No specific laws establish professional independence of staff producing and disseminating GFS; AGPR and provincial AGs have autonomy under law in accounting functions, and a longstanding convention of independence of MOF in providing and publishing fiscal data exists.
- Users can expect GFS on the MOF website to be published within a regular timeframe (two months after the end of the reference quarter); this timeframe is stated on the GDDS website but not on the MOF website.
- There are no professional statistical staff in the MOF and no systematic efforts to promote professionalism of GFS compilers.
- No evidence of political interference in choices of sources and statistical techniques; content and format of statistics are based on IMF recommendations.
- GFS compilers may provide technical advice and respond to misinterpretation of GFS data, but explanatory material is not provided to assist interpretation of published statistics.
- No government access (outside the MOF) to, or ministerial commentary on, GFS before release; this is stated on the GDDS website.
- GFS are identified in the Economic Survey as products of the Budget Wing of MOF; GFS disseminated on the MOF website are not explicitly sourced.
- A major change to source data (use of AGPR/AGs accounting data after an IMF mission in 2000) resulted in substantial revisions to previously published data, but users were not informed about the change or alerted to consequent revisions. The need for advance notice of such changes is acknowledged.

### Methodological soundness — concepts, definitions, and scope
- Compilation of both quarterly and annual GFS follows the analytic framework used by the IMF for its operational work, largely based on A Manual on Government Finance Statistics, 1986 (GFSM 1986). Treatment of privatization receipts resembles GFSM 2001 methodology.
- The published description of external government debt indicates that it includes government guaranteed debt; this treatment is not in line with internationally accepted standards, which require that outstanding government debt be identified separately from debt guaranteed by government. The composition of government debt figures needs clarification and guaranteed debt should be shown separately.
- No specific plans to migrate to the GFSM 2001 framework; authorities are keen to improve existing framework and classification. Presenting existing cash GFS data in the GFSM 2001 framework is an option to explore.
- Scope and coverage:
  - GFS in Pakistan cover the budget sectors for the three levels of government.
  - There are no separate social security funds in Pakistan, but some extra-budgetary funds exist at the central government level and are not covered in GFS; exclusion of extra-budgetary funds results in incomplete coverage of central government transactions.
  - Because the budget sectors in Pakistan cover more than 90 percent of central government transactions and more than 95 percent of general government transactions, data coverage is adequate for fiscal analysis.
  - Coverage of local government extends only to the local government district level; lower-level local government transactions are relatively minor.
  - Data are available on the stocks and flows between budget units (levels of government), and payments and receipts between these units are reconciled by AGPR, allowing consolidated statistics for provincial governments and the whole of general budgetary government.
  - Revenue items are available to approximately the level specified in GFSM 1986.
  - Expenditure data are not provided at the level of detail required by GFSM 1986: expenditure is classified on a mixture of economic and functional classifications, with defense and government administration expenditures shown as separate lines with no economic detail; no other functional detail is provided, and economic detail is far less than required by GFSM 1986.
  - Financing information is classified only to broad counterparty institutional sectors, with little instrument detail.
  - Debt data cover the budget sectors; any borrowing by extra-budgetary units is likely to be done through the budget sectors, so the whole of general government is effectively covered under present arrangements.

*Source: _cr04398 - 3.5 Revision studies (PDF chapter/section).*

### 2.3  Classification/sectorization

### 2.3  Classification/sectorization

### Classification and sector coverage
- Institutional sectors are not explicitly defined; only the budget sectors for each level of government are included in GFS.
- GFS are produced for:
  - central government;
  - provincial governments; and
  - consolidated budgetary government.
- The local government sector is not distinguished from provincial governments; separate data for the local government sector within each province are not available despite initiatives to devolve more power to local government.
- The most significant extra-budgetary fund is ZAKAT, which handles transactions related to Muslim religious observance (providing aid to the poor) but is controlled by the government.
- Pakistan Railways and Pakistan Post are regarded as budget enterprises; their gross operations are not included in the budget, but substantial transfers are made from the budget (especially to Pakistan Railways) for both capital and current purposes and are classified as grants in GFS.
  - The current transfers (the largest component) should be classified as subsidies to public corporations in accordance with GFSM 1986/2001.
- Public corporations undertake some quasi-fiscal activities (e.g., provision of goods and services to the poor at below-market rates). In most cases, additional costs or loss of revenue from these activities are explicitly covered by budget subsidies, so no additional action is required to bring quasi-fiscal activities into the scope of GFS.

### Classification detail and departures from GFSM
- Revenue, lending minus repayments, and financing broadly follow the GFSM 1986 methodology, but significant differences from GFSM 1986 exist for grants and expenditure.
- Revenue and grants:
  - At an aggregate level, transactions are appropriately classified as revenue and grants according to international standards, except:
    - foreign grants received are classified as financing, rather than being included in revenue and grants (1986 standard) or revenue (2001 standard);
    - central government loans to provincial governments are included in provincial government revenue.
  - Tax revenue is distinguished from nontax revenue. At consolidated government level, tax revenue is classified into:
    - (i) taxes on income and profits;
    - (ii) taxes on property;
    - (iii) taxes on goods and services;
    - (iv) taxes on international trade; or
    - (v) other taxes.
  - The category “other taxes” includes motor vehicle taxes that should be classified as taxes on goods and services.
  - Non-tax revenues are defined broadly according to the 1986 GFS methodology.
  - About one quarter of total non-tax revenue is classified as other miscellaneous revenue (September quarter 2003); more detail exists in source data but is not provided in GFS.
- Expenditure:
  - At an aggregate level, transactions are classified as expenditures according to international standards, but the classification detail is far below GFSM 1986 or GFSM 2001 requirements.
  - Expenditures are classified as current or “development and net lending.”
    - Current expenditures are classified into central (“federal”) and provincial, with no further detail for provincial expenditures (which comprise approximately one quarter of the total).
    - For central government expenditures, only very summary detail is provided: “defense and security,” “general administration and services,” and economic items limited to interest payments, grants, and subsidies.
    - Development and net lending distinguish only between capital expenditure—the public sector development program (PSDP)—and net lending (lending minus repayments). No details of capital expenditure by type of nonfinancial asset are provided.
    - A full classification by function of government (COFOG) is not available.
    - Grants include substantial amounts that should be reclassified as subsidies to public corporations (railways).
  - In summary, expenditure classification departs from GFSM 1986 methodology because economic and functional classifications are not clearly distinguished and the level of detail is inadequate.
- Lending minus repayments:
  - Lending minus repayments are appropriately classified.
  - Data for government lending are not provided separately from repayments of loans previously extended by the government.
  - Proceeds from privatization are classified below the line as financing (not in line with GFSM 1986, but consistent with GFSM 2001 treatment of sales of equity as disposal of financial assets).
- Financing:
  - Classified as foreign (external) and domestic.
  - Domestic financing is further classified by sector of counterparty (bank/nonbank), with limited instrument detail for nonbank financing.
  - Proceeds from privatization are part of financing (in accordance with GFSM 2001).
  - Foreign grants are classified as financing, and central government loans received by provincial governments are classified as provincial government revenue; neither treatment is consistent with GFSM 1986 or GFSM 2001.
- Government debt:
  - Data on government debt are classified by currency, type of debt instrument, and type of debt holder, and domestic debt are shown separately from foreign debt.
  - Note: type of debt instrument classification is provided but with limited detail.

*Italic: Source: _cr04398 - 2.3  Classification/sectorization*

### 2.4  Basis for recording

### Valuation and price basis
- Prices used to value transactions reflect actual or expected cash payments as required by GFSM 1986 guidelines.
- Debt liabilities are recorded at face value in accordance with GFSM 1986 (the amount to be repaid at the end of the contract).
- External debt is valued in foreign currency (U.S. dollar).
- GFS requires amounts denominated in foreign currency to be converted to domestic currency using the mid-point exchange rate at the end of each period.

### Recording basis and timing
- Recording is done on a cash basis:
  - All budgetary central government revenues, expenditures and lending minus repayments transactions are recorded based on documentation authorizing transfer of funds to or from the consolidated fund (e.g., checks paid or received, authorization to the SBP to transfer funds to provincial or local governments).
  - In some cases considerable time may elapse between authorization to transfer funds and the actual transfer.
  - Corrections to accounting errors (mis-postings) are entered into the period in which the correction is made, not the period in which the error occurred.
  - Financing data are recorded based on administrative information on incurrence or redemption of liabilities, and SBP accounting on changes to holdings of cash—equivalent or close to the cash basis recommended by GFSM 1986.
- A small statistical discrepancy is expected between overall deficit/surplus and total financing due to timing differences between above-the-line and below-the-line recording (float).
  - In practice, the deficit/surplus is adjusted to agree with financing by adding an additional expenditure item (“unidentified expenditure”), which appears larger and more volatile than expected given nominally cash-basis recording.
- There are no plans to move to an accrual basis of recording.

### Grossing/netting procedures
- Government finance transactions are shown on a gross basis.
- Tax refunds are classified correctly as (negative) revenue transactions but are not recorded in the period in which the original transaction is recorded; this treatment is not in line with internationally accepted standards because refunds should be netted against taxes in the period in which they were originally recorded.

*Italic: Source: _cr04398 - 2.3  Classification/sectorization*

### 3  Accuracy and reliability

### Source data collection and coverage
- Timely and comprehensive budget revenue and expenditure (including lending minus repayments) data for all levels of government are available on a monthly basis from the AGPR/AGs accounting systems.
  - AGPR provides accounting services for central budgetary government; AGs provide equivalent services to provincial governments using the same chart of accounts.
  - Provincial AGs provide monthly accounting data to the Finance Division and AGPR; AGPR uses this data to compile monthly and quarterly GFS tables.
- Financing:
  - Domestic financing data are timely and comprehensive from the SBP for:
    - government accounts at the SBP (daily);
    - government accounts at other banks (monthly);
    - domestic issue and redemption of securities (monthly), except National Savings Schemes (NSS) securities, which are reported monthly by the Central Directorate of National Savings (part of MOF).
  - Data on foreign (external) loans are obtained quarterly from the Economic Affairs Division of MOF (which reconciles with information provided to SBP by foreign loan providers).
- No data are collected from extra-budgetary agencies.

### Source data conformity and timeliness
- Central and provincial budget classification for revenue and expenditure broadly aligns with 1986 GFSM definitions, classifications, time of recording, and valuation.
- For proceeds from privatization, GFSM 2001 methodology is followed.
- Source data for financing conform to GFSM 1986 definitions, classifications, valuation, and time of recording.
- GFS aggregates are not automatically derived from accounting data but could be once the new chart of accounts is in operation.
- Data published on budget outcomes are reconcilable with GFS data in the MOF Economic Survey and on its website.
- Timeliness:
  - Federal budget data are supplied within 15 days of the end of the reference period.
  - Provincial and district government data are sufficiently timely despite lack of full computer automation.
  - Complete, audited final accounts of budgetary government are available about seven months after the end of the reference year.

### Compilation and validation procedures
- GFS are compiled directly from accounting and administrative data; no statistical techniques are applied in compilation.
- Preliminary data are used for initial publication and are replaced by audited data when final accounts are available.
- A statistical discrepancy exists between overall deficit/surplus and total financing; efforts are made to minimize it.
- Main validation is via the “reconciliation” process—vetting of accounting data by line agencies under FMC supervision—and benchmarking preliminary to audited data.
  - Data corrections are posted in the period discovered, not the original posting period; this currently affects annual data only.
  - Deviations between preliminary and final data are usually small but are not highlighted or explained when revised data are disseminated.
- No systematic reconciliation process is followed to ensure GFS are consistent with corresponding data from other sources; the assumption is that financing data are more reliable, and the discrepancy is shown as unidentified expenditure in published data.
- Revision studies:
  - For annual data, revisions mainly result from replacing preliminary with final audited data; no systematic analyses of revisions are regularly performed, although significant deviations are examined.
  - When historical data were revised in 2000 to reflect a new GFS methodology (use of AGPR/AGs data), the revised data were published without explanation.

*Italic: Source: _cr04398 - 2.3  Classification/sectorization*

### 4  Serviceability

### Relevance and practical utility
- Monthly, quarterly, and annual GFS covering budgetary central and general government are used by MOF and IMF to determine, monitor, and evaluate central and general government fiscal policy.
- Monthly GFS are available to authorities and IMF staff though only quarterly and annual data are published.
- GFS fiscal tables are used extensively by authorities and IMF advisors; a user survey indicates wide use by private sector organizations, universities, and the press.
- MOF monitors press comments and economic forum comments but receives almost no direct feedback from statistics users; there is no systematic attempt to ascertain relevance and practical utility of GFS.
- Since Pakistan’s GFS is narrowly defined to meet IMF operational requirements, it is not well suited for more general analytical use.

### Timeliness and periodicity (compliance with GDDS standards)
- Timeliness meets or exceeds GDDS standards:
  - Quarterly budgetary central and general government operation (revenue, expenditure and lending minus repayments, and financing) are disseminated within two months after the end of the reference period. (GDDS recommendation: quarterly central government aggregates within one quarter after the end of the reference period.)
  - Monthly budgetary central and general government domestic debt data are disseminated within two months after the end of the reference period. Quarterly budgetary central and general government foreign (external) debt are disseminated two months after the reference period. (GDDS recommendation: annual central government debt data within one-two quarters after the end of the reference period.)
  - Annual budgetary central and general government data (sum of four quarters) are disseminated two months after the end of the reference period. (GDDS recommendation: comprehensive annual central government data within six-nine months after the end of the reference period.)
- Periodicity meets GDDS standards:
  - Budgetary central and general government aggregates are disseminated quarterly and annually (GDDS recommendation: quarterly and annual central government aggregates).
  - Budgetary central and general government debt statistics (including, not separately identified, government guaranteed debt) are disseminated monthly (domestic debt) and quarterly (external debt). (GDDS recommendation: annual central government debt data, with quarterly data encouraged; GDDS also recommends dissemination of annual government guaranteed debt separately from outstanding government debt.)

*Italic: Source: _cr04398 - 2.3  Classification/sectorization*

### 4.3       Consistency

### 4.3       Consistency

### Internal consistency of GFS transaction data
- GFS transaction data are internally consistent, except for a discrepancy between deficit/surplus (derived from revenue, expenditure, and lending minus repayments) and financing data.
- Two main reasons for the overall deficit/surplus differing from financing:
  - Timing differences: revenue and expenditure are recorded on the basis of documentation authorizing transfer of funds (for example, checks paid or received, authorization to the SBP to transfer funds to provincial or local governments), while financing data more closely reflect actual cash receipts and payments, resulting in a timing difference (float) between the two measures.24
  - Difficulties in obtaining accurate information for revenue and expenditure, and/or financing.
- Current practice:
  - The discrepancy is examined and reconciled as far as possible for each period, but no systematic analysis is done to determine whether the discrepancy cancels out over time or signals more significant consistency problems.
  - The discrepancy (whether positive or negative) is presently included in expenditure as “unidentified expenditure.” It would be more appropriate to explicitly identify it as a separate discrepancy item.
- Government domestic and foreign borrowing data:
  - End-period debt balances are derived by adding the face value25 of new loan contracts incurred during the period, less amortization of existing loan contracts, to the face value of existing loan contracts at the start of a period.
  - Financing data are based on the cash flows associated with issuing and amortization of loan contracts.
  - The change in gross debt from the opening to the closing debt balance would be expected to be similar to, but not the same as, loan financing during the period (excluding any guaranteed debt).26
  - No studies have been carried out to determine whether loan financing is consistent with the change in debt stocks.
- Quarterly and annual data are compiled by adding data for the component periods.

### Consistency over time
- The MOF has compiled a GFS time series covering the period from 1996/97 onward.
- Revenue and expenditure appear to be reasonably consistent as a percentage of GDP over that period, but no studies have been conducted to determine whether the time series is consistent with fiscal policy over that period.
- The MOF does not publish a quarterly GFS time series on its website; the website shows only the cumulative quarterly data for the current year.
- An annual time series is published in the Economic Survey, but with little transaction detail.
- The MOF is aware that it will need to document and explain changes in the time series that will result from any possible changes in the methodology.

### Consistency with other datasets and frameworks
- No GFS cover the operations of the general government in Pakistan. The Statistics Bureau compiles general government data for the national accounts using MOF budgetary general government data.
- Domestic debt statistics are derived from, and therefore are consistent with, monetary and financial statistics.
- GFS transaction data are not reconcilable with either monetary or balance of payments statistics.

### Relevant findings from assessment tables (serviceability/consistency)
- GFS are not reconcilable with monetary and financial statistics, or BOP.
- No studies of whether loan financing is consistent with changes in debt stocks.
- No systematic analysis of whether the deficit/financing discrepancy cancels out over time.

### Practice on revisions and preliminary data (implications for consistency)
- Dissemination follows a regular practice where initial data are preliminary and subsequently replaced by final data.
- Preliminary data are clearly identified in either the table headings or a footnote to the table.
- Corrections owing to data errors are made to data in the period in which the error is discovered, rather than shown as revisions to the original data.
- The annual series published in the Economic Survey show revised estimates, but with no indication of the size of the revisions. There is no published analysis or explanation of revisions.

### Key recommendations related to consistency (from the Recommendations section)
- Examine the statistical discrepancy between the deficit/surplus (net lending/borrowing under GFSM 2001) and financing each quarter,31 and reconcile and/or explain the differences. (This task should be monitored by the Fiscal Monitoring Committee (FMC).)
- Identify the discrepancy between the deficit/surplus and financing explicitly, and consider including it in the financing category.
- Develop processes to compile and disseminate data monthly, with timeliness of one month, as part of the MOF plan.
- Provide the classificatory detail specified in GFSM 2001 in GFS tables, as far as possible, and provide a time series of discrete (noncumulative) quarterly data.

*Source: _cr04398 - 4.3       Consistency*

### Section 40 of the SBP Act specifies that the SBP must prepare for the Federal Government a

### _cr04398 - Section 40 of the SBP Act specifies that the SBP must prepare for the Federal Government a

### Legal mandate and public dissemination
- Section 40 of the SBP Act specifies that the SBP must prepare for the Federal Government a weekly account of the SBP’s Issue Department and Banking Department by notification in the official Gazette, which is available to the general public.
- SBP’s legal mandate to publish the scheduled banks’ statistics is specified in Section 33 of the Banking Companies Ordinance of 1962, which states that the SBP “may publish any information ... in such consolidated form as it thinks fit.”
- Although the responsibility for producing monetary statistics is not explicitly mentioned in the SBP Act, the SBP has compiled and disseminated monetary statistics since its establishment in 1948; this practice is covered implicitly insofar as the SBP requires timely and accurate financial information of the banking system to accomplish its objective to maintain monetary stability in Pakistan.

### Data sharing, compilation responsibilities, and operational arrangements
- Working arrangements: Banking Supervision and Accounts Departments forward data to the Statistics Department; Statistics Department is sole statistical agency to collect, process, and disseminate monetary statistics in Pakistan.
- Reporting by banks:
  - Scheduled banks submit weekly and monthly balance sheet data with additional detailed schedules to the Banking Supervision Department and Statistics Department, respectively.
  - Banking Supervision Department forwards aggregated weekly data to the Statistics Department according to a formal agreement and timetable.
  - Accounts Department provides the Statistics Department with the central bank’s accounting records.
- Outputs and use:
  - Statistics Department compiles and disseminates analytical accounts of the SBP and scheduled banks and compiles a monetary survey in the required format for IMF publication and operational purposes.
  - Economics Policy Department compiles and publishes in the Statistical Bulletin data on reserve and broad money and causative factors for changes in these aggregates.
- Coordination:
  - Regular meetings between departments ensure understanding of data requirements and coordination.
  - Statistics Department seeks views of Research and Economic Policy Departments on relevance and quality of compiled data.

### Confidentiality, legal protections, and enforcement measures
- Confidentiality provisions:
  - SBP Act, Section 53 (1): information related to individual units, persons, or authorities obtained by SBP staff must be treated as confidential (“preserve secrecy”).
  - Penalties for confidentiality violations range from a fine of up to 500 Pakistani rupees (Prs) to imprisonment of up to six months.
  - Banking Companies Ordinance and SBP Staff Regulations (2001) also stipulate confidentiality and nondisclosure with penalties for noncompliance.
- Legal authority to require reporting:
  - SBP Act, Section 36 (3) provides legal authority to collect information from scheduled banks required for compilation of monetary statistics.
  - Reporting requirements: banks required to report weekly and quarterly (where end-of-week dates and end-of-quarter dates fall on different days) balance sheet data to Banking Supervision Department; monthly and half-yearly balance sheet information to Statistics Department.
- Penalties for noncompliance or false reporting:
  - SBP Act, Section 36 (6): any scheduled bank failing to comply liable to pay PRs 100 for each day during which the failure continues.
  - Prudential Regulation 25 (Circular 21 of June 9, 1999): officials of scheduled banks liable to pay PRs 20,000 per case and/or PRs 1,000 per day for the period of the irregularity, and/or disciplinary action.
  - SBP Act, Section 36 (7): legal consequences for willfully or recklessly making false statements in banks’ returns.

### Resources: staff, training, and IT
- Staffing:
  - Fourteen of the 76 staff in the SBP’s Statistics Department are charged with primary responsibility for monetary statistics work.
  - Employment is competitive; all statistical officers, assistant directors, junior joint directors, the joint director, and the director have at least master’s degrees in the field of statistics; three of which hold a Ph.D. degree; some officers have additional qualifications of master’s degrees in economics and/or business administration.
  - Staff rotation is practiced within the department.
- Training:
  - Staff are given opportunity to participate in IMF courses and seminars offered by central banks such as the Bank of China and the Central Bank of the United Arab Emirates.
  - Out of the 14 staff involved in monetary statistics, three staff attended the IMF Monetary and Financial Statistics Course.
  - Nominations to courses are made by a selection committee of the SBP.
  - In view of implementing the Monetary and Financial Statistics Manual (MFSM), staff need more training in statistical methodology and compilation practices.
- Computing and budget:
  - Computer resources for collection and compilation are broadly adequate.
  - Plans contemplated in the department’s budget plan for 2004 to equip every staff with a computer and replace older hardware with Pentium IV computer processing units.
  - There are no budgetary constraints impeding data collection and compilation activities.

### Resource management and accountability
- Business planning:
  - The SBP’s Business Plan (first implemented for fiscal year 2002–2003) governs the statistical program for monetary statistics.
  - The Business Plan describes objectives, operational and developmental projects, and determines human, computing, and financial resources needed for the fiscal year.
  - It outlines key performance indicators and key result areas for individual staff and divisions; these indicators are benchmarks for periodic staff performance reviews.
- Oversight:
  - Efficient allocation of resources assessed in a midyear review by the Audit Department.
  - SBP Governor reviews implementation of quarterly projects by the Statistics Department.
  - Director of the Statistics Department is made accountable to optimally manage allocated resources.

### Quality assurance and methodological processes
- Quality focus:
  - Processes and procedures guided by SBP Vision Statement: “To transform the State Bank of Pakistan into a highly professional, efficient and modern institution which is fully equipped to play a meaningful role in the economic development of Pakistan” (included in the Statistical Bulletin).
  - SBP considers the Data Warehouse a key step to enhance data quality.
  - All data compiled by the Statistics Department are checked by the immediate supervisor of the data producing unit and accuracy is certified prior to publication.
- External audit:
  - SBP Act Sections 43 and 45 require SBP annual balance sheets and related accounts be examined by two auditors appointed by the SBP board of directors.
  - Banking Companies Ordinance Section 35 requires scheduled banks’ balance sheet and profit and loss accounts be audited by a person on the SBP panel of auditors.
- Validation and review procedures:
  - Statistics Department and Banking Supervision Department review source data for accuracy using automatic and visual validation procedures on a bank-by-bank basis.
  - Statistics Department (monthly source data) and Banking Supervision Department (weekly and quarterly returns) consult with reporting banks to verify misclassifications.
  - Economic Policy Department reviews the monetary survey compiled by the Statistics Department for possible adjustments before official release.
  - Banks are provided reporting software by the Statistics Department with built-in validity checks; cross-sectional validation is applied on receipt.
  - The Statistics Department does not utilize formal periodic user surveys to obtain feedback from users on data quality issues.
- Planning and development:
  - Planning and Development Division monitors data quality issues and new developments in monetary and financial statistics and prepares proposals/steps to implement statistical methodology in accordance with international standards.

### Integrity, transparency, and professional standards
- Professionalism and independence:
  - SBP’s responsibility to regulate monetary and credit system and secure monetary stability; autonomy granted in February 1997.
  - No evidence of interference from others, including other government agencies, in compilation or dissemination of statistics.
  - Recruitment in Statistics Department based on aptitude, expertise, and merit via an open competitive process.
  - Performance Management System introduced in July 2002 refined promotion system at SBP.
- Source and method objectivity:
  - Choices of source data (balance sheet records) and statistical techniques are based on measurement objectives and data requirements.
- Right to comment on misuse:
  - Research Department and Corporate and Media Affairs Department (Public Relations Division) maintain records of press articles on SBP data; such articles are sent to relevant departments for comments and public clarifications issued if necessary.
- Transparency of terms, access, and notifications:
  - Public access to SBP Act and Banking Companies Ordinance; SBP directives and regulations on data collection issued as circulars to banks were posted on the SBP website, included in the Statistical Bulletin, and made available to the public after the ROSC mission.
  - Pakistan’s GDDS metadata, hyperlinked to the SBP website, provide processing and dissemination information; IMF’s Data Standards Bulletin Board (DSBB) includes identification of a contact person at the SBP for additional guidance.
  - There is no internal government access to SBP-compiled statistics prior to public release.
  - Monetary statistics disseminated by the SBP are clearly identified as such; when data from other agencies are published, the source is indicated in footnotes.
  - Advance notice and briefings:
    - Statistics Department recently modified classification of assets and liabilities in statements for scheduled banks in line with the International Standard Industrial Classification (ISIC, Rev. 3.1) and conducted briefing sessions with banks.
    - More recently, steps undertaken to improve methodology and data sources, including changes in classification of assets and liabilities to align with ISIC Rev. 3.

### Methodological soundness and formats
- Two data sets produced:
  - Monetary Survey compiled by Statistics Department for IMF presentation; covers accounts of all scheduled banks and the SBP and follows MFSM guidance on sectorization and classification of instruments; based on month-end stocks; data collected electronically (floppy diskettes and e-mails to Statistics Department).
  - Monthly monetary data compiled by Economic Policy Department (based on weekly information from Accounts and Banking Supervision Departments) used for money market management, monitoring and supervision; format not in accordance with recommended format for analytical accounts.
- Dissemination practice at time of mission:
  - Monetary survey compiled by Statistics Department was not yet published in any SBP publication at time of ROSC mission; assessment under DQAF was based on monetary statistics published in the monthly Statistical Bulletin.
- Published frameworks and tables:
  - Analytical framework in Statistical Bulletin reflects concepts based on IMF’s draft Guide to Money and Banking Statistics in International Finance Statistics (December 1984).
  - SBP publishes separate balance sheets for Issue and Banking Departments in conformity with SBP Act, Section 26.
  - Consolidation of SBP sub-balance sheets with aggregated positions of scheduled banks produces monetary survey of banking system, though the term monetary survey is not used in the Statistical Bulletin.
  - Economics Policy Department publishes:
    - Table 2.1, Monetary Statistics: Reserve money (M0), Narrow money (M1), Broad money (M2), and Money multiplier (M2/M0); all positions shown as monthly stock data.
    - Table 2.2, Analysis of Causative Factors for Changes in Monetary Statistics: counterpart components for M0 and M2 shown as flow data and grouped into sectors and instruments including Public sector, Non-government sector, Counterpart funds, Other items (net), and Foreign assets (net).
- Transition to MFSM:
  - Following publication of the Monetary and Financial Statistics Manual (MFSM) in September 2000, SBP intends to adjust procedures and formats for collection, compilation, and dissemination to the new guidelines; present compilation practices compared with MFSM indicate areas where SBP has yet to make changes to comply with the new methodological standard.

*Source: _cr04398 - Section 40 of the SBP Act specifies that the SBP must prepare for the Federal Government a*

### 2.4 are commonly used by the authorities to describe the consolidation of the accounts of the

### _cr04398 - 2.4 are commonly used by the authorities to describe the consolidation of the accounts of the

### Monetary statistics: summary of key findings
- Aggregates in the Statistical Bulletin are not consolidated in an analytical depository corporations framework linking broad money liabilities of the SBP and scheduled banks to their claims on resident and nonresident sectors.
- The Statistics Department produces a monetary survey in a balance-sheet format for monthly IMF reporting and internal SBP use; this survey was not published or posted on the SBP website through end-December 2003 (the reference period of the data ROSC assessment). In February 2004, SBP published Pakistan’s monetary survey for the reporting month of December 2003 on its website.
- Weekly data are used for monitoring; the Statistics Department is collecting month-end data to compile the monetary survey.
- The sectorization and instrument classification applied in Table 2.2 are not consistent with the System of National Accounts 1993 (1993 SNA) or the Monetary and Financial Statistics Manual (MFSM).
- Data in published tables mix month-end conventions: quarter-end data relate to the last working day of the reference month, other end-period data relate to the last Saturday of the reference month.

### Scope and institutional coverage
- Other Depository Corporations (ODC) subsector definition (MFSM): all resident financial corporations (except the central bank) and quasi-corporations mainly engaged in financial intermediation whose liabilities consist of deposits or financial instruments included in broad money (M2).
- Pakistan’s ODC subsector composition (as stated):
  - 40 scheduled banks
  - 8 public sector banks
  - 16 private domestic banks
  - 16 foreign banks
- Definition context: “scheduled bank” per SBP Act, Section 37 — a bank operating in Pakistan with paid-up capital and reserves of not less than PRs 1 billion, as of January 2003, and conducting operations not detrimental to depositors’ interests.
- Coverage includes scheduled banks’ domestic headquarters and domestic branches, and domestic branches of foreign banks operating in Pakistan; balance sheets of banks in liquidation are included.
- Non-Bank Financial Institutions (NBFI) that accept deposits are not included in monetary statistics. NBFIs include development financial institutions, investment banks, leasing companies, mutual funds, housing finance companies, discount houses, venture capital companies, and modaraba companies.
- Deposits held with some depository non-banks represent only about 1 percent of the banks’ total liabilities; exclusion is considered not material for current compilation of monetary aggregates.
- A new NBFC model allows NBFCs to undertake all financial services except banking functions; under MFSM guidance such entities should be classified as other financial intermediaries or financial auxiliaries.

### Classification and sectorization issues
- SBP uses residency to distinguish domestic and external accounts; delineation broadly consistent with Balance of Payments Manual, fifth edition, 1993 SNA, and MFSM.
- MFSM-recommended subsectors for monetary statistics: (1) central bank; (2) other depository corporations; (3) other financial corporations; (4) central government; (5) state and local governments; (6) public nonfinancial corporations; (7) other nonfinancial corporations; (8) other resident sectors (households and nonprofit institutions serving households).
- SBP publications’ sectorization is not fully in accordance with 1993 SNA and MFSM; proper sectorization is crucial for monetary statistics and flow of funds construction.
- Specific classification shortcomings in SBP publications:
  - Table 2.2: “Credit to public sector” does not clearly separate central (federal) government and state and local (provincial) governments. Budgetary support data relate to central government transactions and are not consistently designated as ‘net’. Federal and provincial governments’ commodity operations are aggregated indistinguishably.
  - MFSM recommends separate compilation/dissemination of Net claims on central government and Claims on state and local governments in central bank and depository corporations surveys.
  - Counterpart sectors for sub-positions “Counterpart funds” and “Public sector corporations other than autonomous bodies” are not defined.
  - Table 2.3: some assets (e.g., Bills purchased and discounted) are classified as domestic rather than nonresident claims; SDRs and SDR allocations are not allocated to any sector; loans received from the IMF are included in Other deposits rather than identified as foreign liabilities.
  - Table 2.4 (Weekly Returns for Scheduled Banks): several positions are not sectorized though significant (e.g., Advances other than those to banks represent 99 percent of total advances): (1) Demand liabilities—Others; (2) Time liabilities—Others; (3) Balances—Others; (4) Advances—Others; (5) Bills purchased and discounted; (6) Investment in securities and shares—Others.
- Classification of instruments deviates from MFSM scheme (based on liquidity/legal characteristics). MFSM groups include: (1) Monetary gold and SDRs; (2) Currency and deposits; (3) Securities other than shares; (4) Loans; (5) Shares and other equity; (6) Insurance technical reserves; (7) Financial Derivatives; (8) Other accounts receivable/payable.
- Published balance sheets do not group assets/liabilities into securities other than shares, loans, or shares and other equity. “Other assets” represent about 19 percent of total assets; “Other liabilities” represent about 14 percent of total liabilities, indicating scope for improved classification/sectorization.
- At present, there is no trading in financial derivatives.
- Repo treatment: securities repurchase agreements (repos) are conducted by the SBP and among banks. Currently treated as outright sales in balance sheets, contrary to MFSM recommendation that repos be treated as collateralized loans (i.e., securities remain on asset side of securities provider; a loan should be recorded as an asset of the cash provider and a liability of the cash taker).

### Basis for recording: valuation, accruals, and grossing/netting
- Valuation:
  - MFSM general recommendation: value financial assets/liabilities at market prices or market-price equivalents (fair values); loans are valued at creditors’ outstanding claims (book value: outstanding principal plus accrued interest) without adjustment for expected loan losses.
  - SBP financial statements prepared in accordance with International Accounting Standards. For scheduled banks, applicability of IAS 39 and IAS 40 has been deferred until further notice.
  - SBP Circular No. 20 (Banking Supervision Department, August 4, 2000) requires scheduled banks to adjust all security holdings to market values weekly irrespective of purpose. Sources: Reuters quotations (PKRV) for government securities; Karachi Stock Exchange prices for quoted corporate securities.
  - Monetary and non-monetary gold revalued annually at closing London bullion rates. SDR holdings and SDR liabilities to IMF revalued weekly, monthly, and annually by SBP.
  - Loan portfolios and deposits recorded at book value; loan valuation not adjusted for expected loan losses. Provisions for loan losses recorded in Other liabilities.
  - Foreign-currency denominated stocks and flows translated into Pakistani rupees at market exchange rate prevailing when entered in accounts.
  - In weekly and quarterly statements, unrealized gains/losses from revaluation are recorded in other liabilities on a net basis. MFSM recommends showing net counterpart to valuation changes on liability side as Shares and other equity (valuation adjustment).
  - MFSM recommends compiling data on stocks and each of three flow components: transactions, revaluations, and other changes in volume. In Pakistan, banks do not provide flow information and SBP does not derive transactions from stock data. Counterpart components for M0 and M2 in Table 2.2 reflect total flows (including revaluations and other changes), not transactions alone.
- Accrual basis:
  - MFSM recommends accrual basis compilation; interest due but unpaid should be incorporated into outstanding amounts of financial instruments.
  - SBP balance sheets are compiled on an accrual basis. However, interest accruing on deposits, loans, and securities other than shares is not incorporated into the outstanding amount of the financial asset/liability as MFSM recommends.
  - Scheduled banks recognize interest income as it accrues except for nonperforming debt; past due interest on nonperforming loans is credited to a suspense account. Dividend income recorded as income when received.
- Grossing/netting procedures:
  - Assets and liabilities of SBP and scheduled banks are collected and compiled on a gross basis in line with MFSM general principles.
  - Data presented on a net basis (e.g., “Foreign assets (net),” “Budgetary support (net),” “Net credit to private sector and public sector corporations”) are also shown with underlying gross data.
  - The monetary survey is obtained by cancelling outstanding claims and liabilities between the SBP and scheduled banks and among scheduled banks.

*Source: Excerpt from the IMF country report chapter on Monetary Statistics (content unit: _cr04398).*

### 3.1       Source       data

### 3.1 Source data

### 3.1.1 Source data collection and coverage
- Source data for the central bank analytical accounts: accounting records prepared by the SBP’s Accounts Department on a weekly basis, generated from a newly established chart of accounts providing real-time data disaggregated and consolidated by functions, departments, divisions, and balance sheet categories (assets, liabilities, income, and expenses).
- The Accounts Department prepares a weekly Statement of Affairs of the State Bank of Pakistan (or “accounting statement”), comprising separate balance sheets for the Issue Department and the Banking Department, forwarded in hardcopy to the Statistics Department.
- Weekly accounting statements lack sufficient detail to classify all financial instruments and economic sectors per MFSM; additional information is obtained from the Accounts Department to enable MFSM-compliant classification.
- Source data for scheduled banks: individual returns developed by and collected from the Banking Supervision Department; reporting forms include banks’ balance sheets, income statements, off-balance-sheet items, and supplementary forms.
- The Statistics Department receives from the Banking Supervision Department a two-page hardcopy Weekly Statement of Position of All Scheduled Banks, which does not contain sufficient details to classify all financial instruments and economic sectors per MFSM; more detailed information is obtained from relevant departments.
- Circular No. 2 of March 16, 1993 introduced new monthly report forms for scheduled banks comprising statement of liabilities and assets, classification of credit by borrowers, deposits distributed by deposit holders, and memoranda items. Main advantages of monthly over weekly returns:
  - (1) all scheduled banks report monthly forms directly to the Statistics Department permitting in-depth analyses and cross-checks without delay;
  - (2) reported data refer to the last working day (not the last Saturday) of the reference month, reducing discrepancies with balance of payments month-end data;
  - (3) forms provide data on banks’ foreign liabilities (mainly foreign currency deposits held by nonresidents), not available in weekly returns;
  - (4) brief guidance notes on classification of resident and nonresident accounts and selected instruments.
- Monthly reports largely provide sufficient detail to compile monetary statistics in conformity with MFSM guidelines.
- Despite monthly source data availability, the SBP compiles analytical accounts for publication in the Statistical Bulletin on the basis of weekly returns; monthly data are used only as supplementary data for classifying and sectorizing weekly data.
- For IMF dissemination of analytical accounts of central bank and scheduled banks, the Statistics Department uses source data retrieved from monthly report forms.
- Meetings with financial market participants and the business community to identify new developments are not common.

### 3.1.2 Approximation to MFSM definitions, scope, classifications, valuation, and recording time
- Weekly accounting statement of SBP and weekly returns of scheduled banks are designed primarily for supervisory purposes and do not reasonably approximate MFSM definitions and classifications for central bank survey, other depository corporations survey, and monetary survey; the Statistics Department makes extensive adjustments when compiling monetary authorities accounts and monetary survey for IMF purposes (see Section 3.2.2).
- These adjustments are not currently applied to SBP and scheduled banks balance sheets as disseminated in the Statistical Bulletin.
- Unclassified items:
  - Other assets in SBP’s accounting statement constitute about 20 percent (end-November 2003) of SBP’s total assets.
  - Other liabilities in scheduled banks’ weekly statement constitute 30 percent (end-October 2003) of scheduled banks’ total liabilities.
- In monthly report forms, the item Other private sector for various asset and liability positions indistinguishably includes private nonfinancial corporations and households (including nonprofit institutions serving households). The Statistics Department plans to revise monthly report forms after SBP implements the 1993 SNA and MFSM guidelines.

### 3.1.3 Timeliness of source data
- Weekly balance sheet data of SBP available 10 days after the reference week.
- Statistics Department requires 2-3 days to generate analytical accounts of the central bank from weekly data.
- End-of-fiscal-year accounting data (end-June) of SBP are available for monetary statistics only after SBP financial statements are approved by the Central Board of Directors, which meets every year before the 24th of August; this delay may prevent dissemination within GDDS timeliness specification of 1-2 months.
- To avoid delay, SBP decided that preliminary data will be used in the future for compilation of monetary statistics for the period June–August.
- Aggregated weekly data for scheduled banks are provided by the Banking Supervision Department 10 days after the reference week (Saturday).
- Reporting date for banks’ monthly report forms is the 20th day after the end of the reference month.
- After receipt, the Statistics Department requires about 7-10 days to finalize analytical accounts for scheduled banks.

### 3.2 Statistical techniques

#### 3.2.1 Data collection and processing practices
- Computerized data collection in SBP’s Statistics Department started in March 1993 with the new monthly report forms for scheduled banks; software for data entry and diskette formatting was developed by the Statistics Department.
- Data are reported on diskettes accompanied by completed hardcopy forms and entered into the electronic database; electronic reporting reduces potential processing errors.
- For central bank source data (from Accounts Department) and scheduled banks’ aggregated weekly statements (from Banking Supervision Department), the Statistics Department only receives hard copies despite electronic availability, requiring manual data entry into the database—this can be a source of error and does not support sound statistical techniques.

#### 3.2.2 Data adjustments, transformations, and compilation techniques
- Statistics Department reclassifies source data for IMF-provided monetary authorities’ accounts and monetary survey. Examples:
  - Foreign currency balances of SBP transferred to resident scheduled banks and financial institutions that appear as Other assets in accounting statements (sinking funds) are re-routed to foreign assets because they remain SBP foreign currency reserves.
  - Loans received from the IMF recorded as other deposits in accounting statements are reclassified as foreign liabilities in IMF presentation.
- These adjustment procedures are not applied to SBP data published in the Statistical Bulletin; the Statistics Department is considering compiling and publishing a central bank survey in line with IMF format and MFSM, employing the same adjustments.
- For dissemination of scheduled banks’ balance sheet in the Statistical Bulletin, the Statistics Department adjusts source data from weekly statements using information from monthly report forms. Adjustments include:
  - (1) deduction of capital assigned by resident banks to their branches operating abroad from Balances with banks abroad;
  - (2) sectorization of various asset and liability positions, including identification of government deposits held with scheduled banks;
  - (3) calculation of banks’ foreign liabilities.

### 3.3 Assessment and validation of source data

#### 3.3.1 Routine assessment of source data quality
- Queries on SBP balance sheet data are resolved between Accounts and Statistics Departments.
- Banking Supervision Department reviews weekly data for scheduled banks and discusses questionable data with responding bank.
- Statistics Department uses computerized procedures to monitor accuracy of monthly source data submitted by banks; validation checks built into electronic report forms permit respondents to rectify errors at data entry.
- Formal inconsistencies and out-of-trend values are confirmed with respondents; only the Statistics Department is authorized to make revisions in monthly report forms.
- Source data are analyzed in the context of revisions.

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

#### 3.4.1 Validation against other information
- Monthly source data submitted by banks are checked, if necessary, against other data sources such as the Statistics Department’s half-yearly returns on scheduled banks (contains balance sheet data and detailed information on deposits, advances and bills, and investment in securities).
- Most queries are resolved directly with scheduled banks by telephone or e-mail.

#### 3.4.2 Investigation of statistical discrepancies in intermediate data
- Statistics Department investigates statistical discrepancies (e.g., inter-bank position discrepancies) and their causes.

#### 3.4.3 Investigation of discrepancies in statistical outputs
- Large fluctuations in monetary or credit aggregates are rare and likely occur only after significant methodological changes.
- Economic Policy Department investigates reasons for fluctuations in the time series Other items (net).

### 3.5 Revision studies
- No mechanisms in place for routine revision studies since SBP and scheduled banks data are considered final when first released (see Section 4.4.2).
- Deviations, omissions, and other potential data problems (e.g., erroneous sectorization) are identified and investigated on an ad hoc basis; results are used internally to inform statistical processes.

### 4 Serviceability — relevance, timeliness, consistency, revisions

#### 4.1 Relevance
- Relevance of monetary statistics is largely monitored via interdepartmental SBP meetings; no established consultation process with private sector, academia, or user surveys.
- User requests for additional statistical information are noted and considered in review of statistical work program.
- Director of Statistics Department and division chiefs participate in international and regional meetings and seminars.

#### 4.2 Timeliness and periodicity
- Monthly data for SBP and scheduled banks are disseminated within four weeks after the reference period, consistent with GDDS recommendation for timeliness (one to two months).
- In the case of SBP, end-of-fiscal-year balance sheet in June is disseminated with a three month lag; authorities expect to fulfill GDDS recommendations using provisional data.
- Balance sheet of SBP and consolidated balance sheet of scheduled banks are disseminated monthly, meeting GDDS monthly periodicity and SDDS requirements.

#### 4.3 Consistency

- 4.3.1 Internal dataset consistency:
  - Inconsistencies exist between SBP’s loans and advances to scheduled banks and scheduled banks’ borrowings from SBP; reconciliation of stock with flow data is essentially not possible due to differing classification, aggregation, and sectorization.
- 4.3.2 Time consistency:
  - Monetary statistics are consistent over time; modifications of source data and methodology are indicated in the Statistical Bulletin with brief footnotes (example: introduction of a new report form for banks in December 2001).
- 4.3.3 Consistency with other frameworks:
  - Monetary statistics are, in principle, reconcilable with balance of payments statistics, but discrepancies exist in Net foreign assets (NFA) due to:
    - different exchange rates used to convert foreign-currency stocks into national currency for different end-reference periods (balance of payments uses midpoint exchange rate of end-reference period; monetary statistics use Pakistani rupee amounts converted by banks using exchange rate prevailing at transaction time);
    - different end-reference dates (balance of payments: last day of the month; monetary statistics: last Saturday of the week).
  - Statistics Department regularly checks consistency with balance of payments statistics.
  - Consistency checks between government finance statistics and monetary statistics are not conducted; publicly available data do not permit reconciliation with government finance statistics.

#### 4.4 Revision policy and practice
- 4.4.1 Revisions schedule:
  - No formal revision policy; published data for SBP and scheduled banks are final when first disseminated. Banks may send revisions in their next return, but such revisions are rare and insignificant.
  - Statistical Bulletin shows revisions and explanatory notes only for causative factors explaining changes in broad money liabilities. Users are not informed that SBP does not conduct routine revisions.
- 4.4.2 Identification of preliminary data:
  - Initially published data on reserve money, broad money liabilities, and causative factors for changes of broad money are identified as preliminary.
  - Revised data are disseminated with same level of detail as previously published but are not explicitly identified in the Statistical Bulletin.
- 4.4.3 Public studies of revisions:
  - No public studies and analyses of revisions due to sporadic nature of revisions; internal analyses focus on errors and shortcomings in banks’ reported data.

### 5 Accessibility

#### 5.1 Data presentation, dissemination media, release practices, and availability upon request
- 5.1.1 Presentation and interpretability:
  - Balance sheet of SBP in the Statistical Bulletin is presented in a Federal Government–approved format not designed for monetary statistics purposes; presentation does not provide information on SBP’s foreign liabilities.
  - Loans and advances to the government of former East Pakistan (Bangladesh) are classified as Internal assets and loans from the IMF are included in Other domestic deposits, which does not facilitate interpretation of central bank claims and liabilities.
  - Statistical Bulletin contains charts on yearly development of currency in circulation and monthly growth of scheduled banks’ demand and time deposits.
  - Seasonally adjusted monetary data are not compiled.
- 5.1.2 Dissemination media and formats:
  - Monetary statistics disseminated weekly and monthly in hard copy and on SBP’s website.
  - Data on SBP website are disseminated only as PDF files; user-friendly Excel files are provided on request.
- 5.1.3 Preannounced release schedule:
  - No advance release calendar historically, but internal production schedules provide reasonably predictable release dates for monetary statistics.
  - Footnote: Beginning in February 2004, a publicly announced advance release calendar is available on the SBP website; internal production schedules also provide predictable release dates.
- 5.1.4 Simultaneous availability:
  - Data are released to all users at the same time; monthly Statistical Bulletin publication and Internet posting are simultaneous.
- 5.1.5 Nonpublished subaggregates:
  - Unpublished and nonconfidential subaggregates are made available upon request.

*Source: _cr04398 - 3.1 Source data (Monetary Statistics).*

### 5.2       Metadata       accessibility

### 5.2       Metadata       accessibility

### 5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques
- Metadata on monetary statistics is only provided in summary form through Pakistan’s GDDS web pages posted on the Dissemination Standards Bulletin Board (DSBB) of the IMF.  
- The SBP provides a hyperlink to the GDDS metadata on its website.

### 5.2.2 Levels of detail are adapted to the needs of the intended audience
- The SBP publications, website, and GDDS metadata provide alternative levels of detail on the methodology employed in compiling Pakistan’s monetary statistics.  
- A comprehensive sources and methods document to inform analysts and other users of monetary statistics about how the data are produced exists only as an internal document (Working Manual) circulated within the Statistics and Economic Policy Departments.

### 5.3 Assistance to users

5.3.1 Contact person for each subject field is publicized
- The statistical publications do not identify contact points for enquiries by mail, telephone, facsimile, or by e-mail.  
- Contact points are available on the departmental pages of the SBP website.  
- The information service of the SBP sends queries on monetary statistics from interested users to the concerned divisions of the Statistics Department for answers and clarifications.  
- Users are provided with contact persons of the Statistics Department in the SBP website.

5.3.2 Catalogs of publications, documents, and other services, including information on any charges, are widely available
- A list of statistical publications with charges and their periodicity is included in all SBP publications.  
- The Statistical Bulletin and Annual Report of the SBP along with other publications are also available on the SBP’s website.

### Data Quality Assessment Framework — Summary findings and planned improvements (Monetary Statistics)
- Assessment key: NA = Not Applicable; O = Practice Observed; LO = Practice Largely Observed; LNO = Practice Largely Not Observed; NO = Practice Not Observed; SDDS = Complies with SDDS Criteria.
- 0. Prerequisites of quality
  - 0.1 Legal and institutional environment: X
  - 0.2 Resources: X
    - Comment: Monetary statistics compilers need training in statistical methodology.
  - 0.3 Quality awareness: X
- 1. Integrity
  - 1.1 Professionalism: X
  - 1.2 Transparency: X
  - 1.3 Ethical standards: X
- 2. Methodological soundness
  - 2.1 Concepts and definitions: X
    - Comment: Published monetary and credit aggregates for the SBP and scheduled banks are not presented in a consolidated analytical framework. Published data are not compiled on a end-month basis.  
    - SBP plans: (i) publish an internally consistent analytical framework for monetary statistics as already provided to the IMF (short-term), and (ii) compile the monetary survey using end-month data (short-term). 48
  - 2.2 Scope: X
  - 2.3 Classification/sectorization: X
    - Comment: Instrument classification, sectorization, and recording of repos and financial instruments are not according to MFSM.  
    - SBP plans: fully implement the MFSM guidelines on the sectorization, classification of instruments, and accounting (short-term).
  - 2.4 Basis for recording: X
    - Comment: Accrued interest is not included in the underlying instrument. Gold balances are only revalued on an annual basis.  
    - SBP plans: incorporate accrued interest in the outstanding values of financial instruments (medium-term).
- 3. Accuracy and reliability
  - 3.1 Source data: X
    - Comment: Banks’ weekly data overall do not reasonably approximate instruments and sectors.  
    - SBP plans: compile monetary statistics based on data collected through the monthly report forms rather than the weekly statements provided by the Banking Supervision Department (short-term). 49
  - 3.2 Statistical techniques: X
    - Comment: Weekly data on SBP and banks are entered manually into the database, which can be a source of errors.  
    - SBP has initiated a data warehouse project to improve data capturing procedures for SBP and banks’ weekly balance sheet data (short-term).
  - 3.3 Assessment and validation of source data: X
  - 3.4 Assessment and validation of intermediate data and statistical outputs: X
  - 3.5 Revision studies: X
- 4. Serviceability
  - 4.1 Relevance: X
    - Comment: No formal procedures for consultations with users, such as periodic meetings or user surveys, to solicit feedback.  
    - SBP plans: formally seek feedback from users on the relevance and usefulness of data on a regular basis (medium-term).
  - 4.2 Timeliness and periodicity: X
    - Comment: SBP data for end-fiscal year (June) are disseminated with a three-month lag.  
    - SBP has decided to use provisional data to improve the timeliness of SBP data for end-fiscal year (short-term).
  - 4.3 Consistency: X
    - Comment: Large discrepancies exist in the positions between the SBP and scheduled banks. Monetary data are not reconcilable with government finance statistics.  
    - SBP plans: improve data consistency between the central bank and scheduled banks, and between monetary statistics and government finance statistics (short-term).
  - 4.4 Revision policy and practice: X
    - Comment: There is no formal revision policy. Revisions are infrequent and made when needed. Users are not informed of this revision policy for monetary statistics.  
    - SBP plans: inform users of its revision policy and practice, and include documentation of revisions in the statistical publications (short-term).
- 5. Accessibility
  - 5.1 Data accessibility: X
    - Comment: Dissemination of foreign liability data and improved instrument classification and sectorization required to facilitate sound analysis.  
    - SBP plans: publish monetary data in a clear manner to facilitate accurate interpretation and analysis (short-term). 50
  - 5.2 Metadata accessibility: X
    - Comment: Metadata are available only on IMF’s Dissemination Standards Bulletin Board.  
    - SBP plans: publish comprehensive metadata for monetary statistics (short-term).
  - 5.3 Assistance to users: X
    - Comment: Contact points are not identified in the Statistical Bulletin.  
    - SBP plans: identify contact points for enquiries by mail, telephone, facsimile, or by e-mail (short-term). 51

Notes in the assessment tables:
- 48 Beginning in June 2004, the SBP disseminates monetary statistics based to a large extent on the MFSM on the web site and the Statistical Bulletin.
- 49 Beginning in June 2004, the SBP publishes monetary statistics based on month-end data.
- 50 Data disseminated from June 2004 onwards follows more closely the MFSM recommended format.
- 51 This information is being provided since early 2004.

### Recommendations for Monetary Statistics (selected, verbatim)
- Prerequisites of data
  - Continue to increase resources dedicated to training in monetary statistics methodology and compilation methods, including international statistical guidelines.
- Methodological soundness
  - For publication in the Statistical Bulletin, compile and disseminate an internally consistent analytical framework that would consolidate the analytical accounts of the SBP with the scheduled banks. Such a framework (monetary survey) is already being provided to the IMF for surveillance and publication purposes. (In February 2004, the SBP disseminated on its website a monetary survey broadly along the MFSM lines).
  - Use month-end data for compiling the analytical accounts of the SBP and scheduled banks.
  - Implement the MFSM guidelines on the sectorization of institutional units, classification of financial instruments, and the recording of securities repurchase agreements (repos) and financial derivatives.
  - Incorporate accrued interest in the outstanding values of financial instruments, as recommended in the MFSM.
- Accuracy and reliability
  - Compile monetary statistics based on the data collected through the monthly report forms rather than the weekly statements provided by the Banking Supervision Department.
  - Improve the data capturing procedures for the SBP and banks’ weekly balance sheet information by using electronic means to enter these data into the database. (This recommendation is being gradually implemented with the ongoing progress of putting into service the data warehouse project of the SBP.)
- Serviceability
  - Improve the data consistency between the SBP and scheduled banks.
  - Establish a regular mechanism to verify the consistency between monetary and government finance statistics.
  - Inform users of SBP’s revision policy and practice and include a documentation of revisions in the statistical publications.
- Accessibility
  - Publish monetary data in a clear manner, especially by improving the clarity of the tables, to facilitate accurate interpretation and analysis (e.g., publish foreign liabilities and disaggregate the nonpublic sector in line with the MFSM guidelines).
  - Publish comprehensive metadata on the concepts and methodology used to collect and compile monetary statistics.
  - Identify contact points for enquiries by mail, telephone, facsimile, or by e-mail.

*Source: _cr04398 - 5.2       Metadata       accessibility*

### 0.3       Quality       awareness

### 0.3       Quality       awareness

### 0.3.1 Processes to focus on quality
- Standing Committee on Balance of Payments and BOPD continuously monitor quality and undertake verification initiatives, including:
  - redesigning the forms for obtaining data of Outstanding Export Bills,
  - direct connectivity to the MOF Data Management FAS system with SBP,
  - monthly reconciliation of official debt service payments with the MOF,
  - monthly reconciliation of reserves with the Accounts and Exchange Debt Management Department,
  - cross-checking of ITRS coding with the respective forms.
- Export Reconciliation Committee verifies trade data compiled by the Federal Bureau of Statistics (customs records) and the SBP (exchange records).

### 0.3.2 Processes to monitor collection, processing, and dissemination quality
- Management of SBP and the Statistics Department regularly monitor BOP data quality and implement procedures for:
  - data entry checks,
  - mismatching checks,
  - out-of-range records,
  - validation listings.
- Supervisor-certification mechanism: immediate supervisor of the data compiler must thoroughly check accuracy and provide a certificate.
- Code guides and FAQs for electronic data collection are available on the SBP website.
- Banks receive regular guidance for coding complex transactions.

### 0.3.3 Processes to handle quality trade-offs and planning for needs
- Management reviews trade-offs between accuracy and timeliness on an ongoing basis, setting cutoff dates for producing data of adequate quality; various cutoff dates are used depending on the time series.
- Planning and Development Division in the Statistics Department:
  - reviews the existing compilation framework and publications,
  - provides guidelines for implementation of international standards and emerging needs.
- No formal user surveys exist; SBP publications invite comments in the preface and feedback—mainly from researchers—is considered.

### Integrity (selected findings)
- Professionalism:
  - SBP compiles statistics impartially; government officials do not comment on or regulate releases, and no revisions have been requested by government officials in the past.
  - Information collected for statistical purposes may not be used by other agencies for other purposes.
  - Staff hiring follows SBP appointment policies with public advertisements and competitive selection.
  - SBP promotes research, provides training, incentives for staff papers, and scholarships for further education.
- Transparency:
  - Pakistan’s Balance of Payments annual publication contains methodology; ITRS Code guide (BPM5 conceptual framework) is available on the SBP website; methodology is available on the IMF website in compliance with GDDS.
  - Provisional data are released on the SBP website without prior government access; final data are referred to the MOF prior to release only for reconciliation of official foreign economic assistance and associated debt service.
  - SBP requests attribution when its statistics are used or reproduced.
  - Advance notice and commentary are provided on revisions and methodological changes to data users and suppliers (example: ISIC implementation for ITRS for BPM5 reporting).
- Ethical standards:
  - Staff behavior guidelines are in the State Bank of Pakistan Staff Regulations (2001), which must be witnessed and signed by all staff members.
- Media handling:
  - A media advisor in the Corporate Affairs Department handles erroneous media interpretations and issues corrective notices.

### Methodological soundness (selected findings)
- Concepts and definitions:
  - Framework generally follows BPM5. Presentation format still follows BPM4 on SBP website/publications, but standard components compiled on BPM5 basis since July 2003 using mapping/bridging with BPM4.
  - SBP plans to aggregate/present monthly BOP data in BPM5 format beginning in January 2004 for internal purposes and in July 2004 for public dissemination.
- Scope and residency:
  - Definitions generally follow BPM5 but exceptions exist: enterprises in EPZs and offshore banks are regarded as nonresidents; Pakistani nationals residing abroad with a dwelling in Pakistan are regarded as residents—this deviates from BPM5 residency criteria.
- Classification/sectorization:
  - Anticipating BPM5 official implementation date January 2004; classification systems include current account, capital account, financial account with BPM5 sectors: monetary authorities (SBP and SBPBSC), general government (federal and provincial ministries, including defense), deposit money banks (commercial banks), and other sectors (nonbank financial institutions, enterprises, individuals).
  - Goods: includes goods for processing by EPZ enterprises; goods procured in ports by carriers and repair of goods are compiled.
  - Services: includes transportation, travel, and other services (communications, construction, insurance, financial services, computer and information services, royalties and license fees, other business services). Transportation includes earnings/expenditures of Pakistani airlines, shipping, and tanker companies; freight and passenger fare earnings of foreign carriers are included.
  - Income: includes compensation of employees, direct investment income, reinvested earnings (comprise amounts issued as bonus shares), portfolio investment income, other investment income; counterpart to exports of crude oil under production sharing agreements is included.
  - Current transfers: includes taxes and duties, cash grants to government, workers’ remittances, counterpart to inflows/outflows of foreign currency accounts of residents, donations, other transfers.
  - Capital transfers: includes migrants’ transfers and official aid received as capital goods.
  - Financial account: direct and portfolio investment definitions follow BPM5.
  - Other investment: includes trade credit assets/liabilities, loans (including Use of Fund credit), currency and deposits, other accounts receivable/payable; sectorized between monetary authorities, general government, banks, other sectors.
  - Currency and deposits: includes foreign assets/liabilities of residents under foreign currency accounts and counterparts of private transfers settled through exchange companies and land-borne trade to Afghanistan.
  - Reserves: includes SDRs, reserve position in the Fund, foreign exchange reserve assets (currency and deposits, securities); balances with banks of commercial banks under SBP authority are included.
- Main deviations and classification issues:
  - Goods account: EPZ enterprises are treated as nonresidents (outside tariff wall); goods exports exclude EPZ exports; goods leaving domestic economy into EPZs are included and classified as goods for processing; imports by EPZ enterprises are excluded; goods entering domestic economy are considered goods imports and classified as goods for processing.
  - Income account: counterpart to crude oil exports under production-sharing agreements included in income account; conceptually this mixes direct investment income and recuperation of capital that should be in financial account.
  - Reinvestment of earnings: inflow of reinvested earnings is underestimated—limited to earnings intended to be paid as bonus shares; other retained earnings allocated to reserves accounts are excluded.
  - Loans: loans guaranteed by government are classified in the general government sector; classification should depend on beneficiary status.

### Basis for recording
- Valuation: transactions in ITRS are settled at market prices; official grants in kind included at prices determined by foreign donors; equity contributions in kind by foreign direct investors valued at prices determined by customs authorities.
- Recording: balance of payments compiled on an accrual basis; adjustments made for timing differences where feasible; BOPD adjusts trade transactions for accrual/timing.
- Grossing/netting: current account items recorded on a gross basis; financial items recorded on a net basis separately for assets and liabilities; recording follows BPM5 guidelines.

### Accuracy and reliability — source data
- Main source data:
  - Returns of banks, enterprise reports, government-provided information.
  - Under ITRS, monthly foreign exchange returns from authorized dealers (ADs) are main source for goods, services, current transfers, direct investment, portfolio investment, debt servicing, other items; these data are compiled on a cash basis.
  - A Code Guide for coding/classification/preparation of returns is provided to all ADs; software is provided to banks for electronic submission.
  - There are at present more than 1,700 ADs operating in Pakistan.
  - ITRS provides complete information on cash inflows/outflows between residents and nonresidents settled through the banking system, broken down by currency and country; transactions classified on BPM5 guidelines.
  - Under the ITRS, data are received through floppy disks from banks’ head offices.
- Other source data:
  - Monthly reports from shipping and airline companies for goods procured in ports, transportation services, and adjustments to f.o.b. basis; reports also contain stock data on foreign assets/liabilities of these companies.
  - Monthly reports from Pakistan’s diplomatic missions for flows and stocks of foreign assets/liabilities.
  - Statements on outstanding export bills from banks to reconcile customs export data with ITRS-derived goods exports; contra-entries included in trade credits.
  - Direct investment flows from ITRS; capital equity contributions in kind and reinvestment of earnings from annual enterprise survey of foreign investment in Pakistan (581 enterprises in 2001).
  - Equity capital of foreign banks from SBP Abstract of Accounts.
  - Portfolio investment flows from ITRS; data on official Special U.S. dollar bonds, euro bonds, government securities from SBP Abstract of Accounts and administrative records.
  - Trade credits from Statement of Outstanding Export Bills; bills discounted by banks from Statement of Affairs (Banking Supervision Department).
  - Pre-shipment and post-shipment export loans and imports from ITRS.
  - Inflows of loans to private sector from agreements provided to Exchange Policy Department; an annual SBP survey of loans received by private sector settled through banks was initiated in July/August 2002.
  - Exchange Entitlement Certificates (EECs) data used; EECs entitle loan recipients to obtain foreign exchange when servicing loans is due.
  - Long-term loans received by government from Economic Affairs Division, including military debt; guaranteed loans from Finance Division.
  - Cash disbursements on loans derived from ITRS and SBP Abstract of Accounts.
  - Inflows of foreign economic assistance, official debt servicing, inflows/outflows on loans received obtained from MOF; private sector loan data from Exchange Policy Department and ITRS.
- Processing and cross-checking:
  - Software developed for computerized cross-checking of enterprise data with cash flows through banks and the SBP.
  - All data processed through a computerized system for timely and accurate dissemination.
- Source data alignment with definitions and recording:
  - Goods from ITRS are on an f.o.b. or c.i.f. basis; airline/shipping/tanker reports used to adjust exports to f.o.b.
  - Imports from ITRS are on a c.i.f. basis; a c.i.f./f.o.b. ratio of 8 percent is used to adjust goods imports to an f.o.b. basis.
  - Coverage adjustments for imports under non-repatriable investment scheme, sales of duty free shops, imports under foreign economic assistance, military imports, exports of crude oil, land-borne trade, oil imports under the Saudi oil facility.
  - Travel payments were underreported after foreign exchange liberalization because money changers were excluded; from April 2003, inclusion of exchange companies significantly improved travel data.
  - Contra entries for reinvested earnings, direct payments of loans of public enterprises, and export of oil are adjusted for coverage and accrual.
  - Separate surveys collected on inflows/outflows to improve loan coverage; direct financing and utilization of loans for PIA aircraft included.
  - Deposits in foreign currency accounts are reclassified from banks to foreign currency deposits of other sectors.
  - Reserves reduced by foreign exchange assets that commercial banks maintain as reserve requirements with SBP; those assets classified as currency and deposits of banks.
- Timeliness:
  - All source data are timely and received on schedule; in case of delays, SBP staff issue reminders and follow-ups via telephone, fax, letters, or e-mails.

*Source: _cr04398 - 0.3       Quality       awareness*

### 3.2       Statistical       techniques

### 3.2 Statistical techniques

### 3.2.1 Data compilation employs sound statistical techniques
- Software checks for completeness, consistency of opening and closing balances, coding validations, period of reporting, etc., of the primary data.
- For ITRS, complete software with online search of codes is provided to automatically perform routine checks and highlight errors at punching level.
- Data are not accepted unless all checks are validated.
- New forms are designed by the BOPD after research on availability of information, frequency of reporting, clarity of items, unit of currency, etc., and items are matched with balance of payments standard components.
- Respondents are briefed through workshops or on a case-by-case basis; most current forms have been in use for many years.
- Respondents are informed of any amendments or additions prior to finalization of revised report forms; necessary training is provided to all potential respondents.

### 3.2.2 Other statistical procedures employ sound statistical techniques
- Moving average percentages of imports by routes compiled by the Federal Bureau of Statistics (ten years data) are used for the distribution of freight on imports by sea, air, and other modes of transportation.

---

### 3.3 Assessment and validation of source data

### 3.3.1 Routine assessment of source data
- Verification entails consistency checks within survey forms and linking data at subsequent levels.
- Accuracy ensured by enforcing that stocks reported at the end of the period equal stocks reported at the beginning plus transactions and other changes; most SBP report forms are based on a closed system.
- For banks, electronic reporting is required on software provided by the SBP with built-in validation checks allowing error rectification at data entry.
- High-value or unusual transactions are investigated and confirmed with respondents.
- All transactions reported by banks are backed by forms filled out by remitter or beneficiary.
- Banks must submit hard copies of foreign exchange returns to foreign offices of the SBP/BSC for validation per balance of payments code guide.
- When revisions are required due to structural or policy changes, source data are analyzed; report forms and population of respondents are revised and/or enlarged as appropriate.

---

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

### 3.4.1 Validation against other information
- Items with second-source data are regularly validated.
- Exports and imports reconciled with Federal Bureau of Statistics.
- Debt service data of official, guaranteed, and commercial loans cross-checked with MOF.
- Private sector debt validated with annual survey of companies through banks.
- Official reserves verified with Accounts and Exchange Debt Management Departments.
- Stocks of foreign currency accounts reported by banks and their treasury departments are cross-matched.
- Financial press monitored to identify high-value direct investments and other transactions for checking.

### 3.4.2 Assessment and investigation of statistical discrepancies
- Bank data for supervisory and monetary policy reconciled with balance of payments data reported by banks.
- In the ITRS, high values or unusual items are investigated.
- A form introduced to assess export bills, linking new export bills and realization of existing bills during the month.
- Stocks of foreign accounts cross-checked between head-office and main branch reports.
- Payments advices for official debt servicing cross-checked by computer with bank and SBP transactions.
- Monthly discrepancies are referred to the MOF for rectifications.
- Private sector outflows settled through banks and direct payments from funds held abroad cross-checked with agreements and repayment schedules in the Exchange Policy Department and the Statistics Department.

### 3.4.3 Investigation of discrepancies and indicators of problems
- Differences in exports between balance of payments and customs data increased significantly during the late nineties; studies with banks and customs led to introduction of a revised form in July 2000, eliminating differences between the two datasets.
- Reporting of inflows and outflows in the new foreign currency accounts with the banking system was not required (leading to underreporting); inclusion of external transactions settled through these accounts became part of the ITRS in April 2003.
- In 2002, banks were authorized to utilize balances held in the new foreign currency for settlement of various external transactions (pre-shipment and post-shipment export credits, import credits, swaps, placements in Pakistan and abroad, etc.), causing mismatches in various balance of payments items.
- In early 2003, the BOPD introduced a new statement on the utilization of these deposits; this statement forms the basis for a number of adjustments in the monthly balance of payments statement.

---

### 3.5 Revision studies

### 3.5.1 Studies and analyses of revisions
- Studies and analyses of revisions are not conducted.
- Large revisions are explained when they occur.
- The summary table notes: "Revision studies are not carried out. Studies of revisions will be carried out in the near future."

---

### 4 Serviceability

### 4.1 Relevance
- Main users: MOF, SBP, and other official agencies; needs met through the Balance of Payments Standing Committee.
- Ad-hoc user requests (often students) occur; no users’ surveys.
- Data users can address questions electronically to bop.query@sbp.org.pk.

### 4.2 Timeliness and periodicity
- Quarterly balance of payments statistics are disseminated within one quarter after the reference period, thus meeting SDDS requirements.
- Annual balance of payments statistics are disseminated within three months after the end of the reference period.
- Balance of payments statistics are disseminated within three months after the end of the reference period, in accordance with SDDS requirements.

### 4.3 Consistency
- Concepts, definitions, and classifications for producing quarterly and annual balance of payments statistics are identical.
- Annual data are the sum of quarterly data.
- Available data are consistent over time.
- Changeover from BPM4 to BPM5 will be implemented provided earlier period information permits maintenance of consistency; footnotes will indicate breaks in series or introduction of new methodology.
- BOPD adjusts data from various sources to reconcile similar data, including:
  - reconciliation of goods account derived from ITRS and customs returns;
  - reconciliation of external debt flows with similar data from the ITRS;
  - incorporation of contra-entries for inflows and outflows in foreign currency accounts, exports of crude oil, and travel flows settled through exchange companies.
- BOP data are reconciled with monetary statistics but not with government finance data; summary table notes: "Data are not reconciled with GFS. Efforts will be made to reconcile data with GFS."

### 4.4 Revision policy and practice
- Revision practice is predetermined and made known to the public.
- Provisional balance of payments data are compiled monthly for supervisory and monetary policy management.
- Data are revised when final data are received.
- Revised and final data are released quarterly.
- Changes are highlighted for major differences between provisional and final data.
- Revised data are always disseminated with the same level of detail.
- Preliminary data are clearly identified; users are alerted that initially published data are subject to revision for up to one year.
- There are no studies or analysis of revisions; revised data are identified in the database accessible to users.

---

### 5 Accessibility

### 5.1 Data accessibility
- Balance of payments data presented clearly to facilitate interpretation and comparisons.
- Four quarterly periods are presented in both rupees and U.S. dollars in Pakistan’s Balance of Payments July 2001-June 2002.
- Regional data reflecting Pakistan’s balance of payments with important partner countries are presented.
- Four annual periods are presented, in U.S. dollars, in the Statistical Bulletin (no accompanying text, tables, or definitions).
- Balance of payments data are published in Pakistan’s Balance of Payments and in the Statistical Bulletin of the SBP, and also available on the SBP website.
- Data will be presented in the format recommended in the BPM5 from January 2004 onward.
- Balance of payments statistics are released in accordance with a well-established schedule known by users; no formal preannounced schedule is publicly available.
- Data are released simultaneously to all interested parties by posting on the website; analytical tables are sent to the MOF for publication in the Economic Survey.
- Unpublished but nonconfidential data or tabulations are made available upon request, but their availability is not publicized.

### 5.2 Metadata accessibility
- Description of Pakistan’s balance of payments methodology and data sources is published in Pakistan’s Balance of Payments and posted on the website; it is also published in the IMF’s Balance of Payments Statistics Yearbook and hyperlinked to the DSBB.
- No subsets of metadata are available; less detailed information on the balance of payments is available in the GDDS metadata on the DSBB.

### 5.3 Assistance to users
- An e-mail address (www.sbp.org.pk) is included in the Statistical Bulletin for requests for information, but no specific contact information for balance of payments data is provided; queries are transmitted to the relevant department.
- A contact person is identified on the DSBB.
- A list of SBP publications is available in the Statistical Bulletin and in the Balance of Payments July 2001-June 2002.
- Most publications are available at a charge; some are complementary or available on the SBP website.

---

*Source: _cr04398 - 3.2 Statistical techniques (Compiling Agency: State Bank of Pakistan) — Balance of Payments Statistics.*

### 5.3  Assistance  to  users

### 5.3  Assistance  to  users

### Assistance summary
- An e-mail address through which enquiries can be made is available.
- Assistance to users will be improved by providing specific contact information for balance of payments statistics.

### Methodological soundness
- Compile data on the external transactions of enterprises operating in the export processing zones and offshore banks and treat them as residents of Pakistan.
- Disaggregate the counterpart of exports of crude oil under product sharing agreements into (i) an income component, to be classified in the direct investment income account, and (ii) a component for the repatriation of direct investment capital, to be classified in the account for foreign direct investment in Pakistan of the financial account.
- Strengthen the procedures to estimate reinvested earnings to include, in addition to earnings capitalized as bonus shares, all earnings capitalized in various other reserve accounts.
- Reclassify government-guaranteed loans from the general government sector to the sector of the beneficiary of the loan.

### Accuracy and reliability
- Conduct revision studies.

### Serviceability
- Develop and disseminate an advance release calendar.
- Establish regular procedures to verify consistency between BOP and GFS.

### Accessibility
- Improve assistance to users by providing more detailed contact information in the publications and on the website.

*The SBP reports significant advances in the implementation of the recommendations during 2004.*

### 0.1 Legal and institutional

### 0.1 Legal and institutional

### Legal and institutional environment — summary of framework and indicators
- The environment is supportive of statistics.
- Elements and indicators focus on "pointers to quality" across five dimensions.

- Legal/institutional responsibility and coordination
  - 0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified.
  - 0.1.2 Data sharing and coordination among data producing agencies are adequate.
  - 0.1.3 Respondents' data are to be kept confidential and used for statistical purposes only.
  - 0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response.

- Resources
  - 0.2 Resources – Resources are commensurate with needs of statistical programs.
  - 0.2.1 Staff, financial, and computing resources are commensurate with statistical programs.
  - 0.2.2 Measures to ensure efficient use of resources are implemented.

- Quality awareness and management
  - 0.3 Quality awareness – Quality is a cornerstone of statistical work.
  - 0.3.1 Processes are in place to focus on quality.
  - 0.3.2 Processes are in place to monitor the quality of the collection, processing, and dissemination of statistics.
  - 0.3.3 Processes are in place to deal with quality considerations, including tradeoffs within quality, and to guide planning for existing and emerging needs.

### Integrity — professionalism, transparency, and ethics
- Core principle: Firm adherence to the principle of objectivity in the collection, compilation, and dissemination of statistics.
- Professionalism
  - 1.1 Professionalism – Statistical policies and practices are guided by professional principles.
  - 1.1.1 Statistics are compiled on an impartial basis.
  - 1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations.
  - 1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics.
- Transparency
  - 1.2 Transparency – Statistical policies and practices are transparent.
  - 1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public.
  - 1.2.2 Internal governmental access to statistics prior to their release is publicly identified.
  - 1.2.3 Products of statistical agencies/units are clearly identified as such.
  - 1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques.
- Ethical standards
  - 1.3 Ethical standards – Policies and practices are guided by ethical standards.
  - 1.3.1 Guidelines for staff behavior are in place and are well known to the staff.

### Methodological soundness — adherence to international standards
- Overarching statement: The conceptual basis for the statistics follows international standards, guidelines, and agreed practices.
- Concepts and definitions
  - 2.1 Concepts and definitions – Concepts and definitions used are in accord with internationally accepted statistical frameworks.
  - 2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices: see dataset-specific framework.
- Scope
  - 2.2 Scope – The scope is in accord with internationally accepted standards, guidelines, or good practices.
  - 2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices: see dataset-specific framework.
- Classification/sectorization
  - 2.3 Classification/sectorization – Classification and sectorization systems are in accord with internationally accepted standards, guidelines, or good practices.
  - 2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices: see dataset-specific framework.
- Basis for recording
  - 2.4 Basis for recording – Flows and stocks are valued and recorded according to internationally accepted standards, guidelines, or good practices.
  - 2.4.1 Market prices are used to value flows and stocks.
  - 2.4.2 Recording is done on an accrual basis.
  - 2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices.

### Accuracy and reliability — source data, techniques, validation, and revisions
- Objective: Source data and compilation techniques are sound, and disseminated data sufficiently portray reality.
- Source data
  - 3.1 Source data – Source data available provide an adequate basis to compile statistics.
  - 3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions.
  - 3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required.
  - 3.1.3 Source data are timely.
- Statistical techniques
  - 3.2 Statistical techniques – Statistical techniques employed conform to sound statistical procedures.
  - 3.2.1 Data compilation employs sound statistical techniques.
  - 3.2.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques.
- Assessment and validation
  - 3.3 Assessment and validation of source data – Source data are regularly assessed and validated.
  - 3.3.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and non-sampling error; the results of the assessments are monitored and made available to guide planning.
  - 3.4 Assessment and validation of intermediate data and statistical outputs – Intermediate results and statistical outputs are regularly assessed and validated.
  - 3.4.1 Main intermediate data are validated against other information where applicable.
  - 3.4.2 Statistical discrepancies in intermediate data are assessed and investigated.
  - 3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated.
- Revision studies
  - 3.5 Revision studies – Revisions, as a gauge of reliability, are tracked and mined for the information they may provide.
  - 3.5.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes.

### Serviceability — relevance, timeliness, consistency, and revision policy
- Goal: Statistics are relevant, timely, consistent, and follow a predictable revisions policy.
- Relevance
  - 4.1 Relevance – Statistics cover relevant information on the subject field.
  - 4.1.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored.
- Timeliness and periodicity
  - 4.2 Timeliness and periodicity – Timeliness and periodicity follow internationally accepted dissemination standards.
  - 4.2.1 Timeliness follows dissemination standards.
  - 4.2.2 Periodicity follows dissemination standards.
- Consistency
  - 4.3 Consistency – Statistics are consistent within the dataset, over time, and with major datasets.
  - 4.3.1 Statistics are consistent within the dataset (e.g., accounting identities observed).
  - 4.3.2 Statistics are consistent or reconcilable over a reasonable period of time.
  - 4.3.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks.
- Revision policy and practice
  - 4.4 Revision policy and practice – Data revisions follow a regular and publicized procedure.
  - 4.4.1 Revisions follow a regular, well-established and transparent schedule.
  - 4.4.2 Preliminary data are clearly identified.
  - 4.4.3 Studies and analyses of revisions are made public.

### Accessibility — data, metadata, and user assistance
- Principle: Clear data and metadata are easily available and assistance to users is adequate.
- Data accessibility
  - 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 formats are adequate.
  - 5.1.3 Statistics are released on the preannounced schedule.
  - 5.1.4 Statistics are made available to all users at the same time.
  - 5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request.
- Metadata accessibility
  - 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.
- Assistance to users
  - 5.3 Assistance to users – Prompt and knowledgeable support service is available.
  - 5.3.1 Contact person for each subject field is publicized.
  - 5.3.2 Catalogues of publications, documents, and other services, including information on any charges, are widely available.

### Users’ views — Pakistan user survey (mission visit December 1-16, 2003)
- Survey design and response
  - The mission conducted an informal survey during December 1-16, 2003 of academics, the media, business associations, banks, public sector agencies, international organizations, and embassies.
  - Fifty-two of the 112 contacted users responded. Response rate: 46 percent.
  - Of respondents: 35 percent private sector; 25 percent embassies and international organizations; 17 percent public sector agencies; 19 percent academics; 4 percent media.

- Aggregate quality perception
  - On a five point scale (1 poor, 5 excellent) respondents rated overall quality of official statistics 3.16 (slightly above average).
  - Overall weighted average rating shown by category: 3.29 3.50 3.60 2.00 3.57 2.60 3.00 and overall 3.16 (table line: Overall weighted average rating 3.29 3.50 3.60 2.00 3.57 2.60 3.00 3.16).
  - Users from international organizations, banks, and public sector agencies generally consider official statistics above average; users in academia and the press rate them below average or poor.

- Key positive points from respondents
  - A large margin of users were particularly satisfied with the ease in accessing official data.
  - Users were also satisfied to a lesser degree with coverage, timeliness, and access to explanatory notes.
  - Users considered Pakistan’s statistics about the same as those of other countries in the region.

- Main criticisms and suggested improvements (summarized from free-text comments)
  - Coverage gaps and detail needs
    - National accounts and production indices are severely under reported and under covered.
    - Fiscal data should contain greater breakdown on budgetary accounts and debt servicing.
    - Monthly exports and imports data lack sufficient detail (variety/quality).
    - Economic data at provincial level insufficient (e.g., provincial contribution to GDP, total and by sectors, and labor market).
    - Social and environmental indicators insufficient at provincial and national level.
    - Coverage of health and education data is sporadic.
    - Insufficient coverage of services sector, micro-economic trends, production, and price trends.
    - Data coverage for nonbank financial institutions, including leasing companies, insurance companies, etc., needs improvement.
  - Periodicity and timeliness weaknesses
    - Periodicity should be improved for most statistics; many series are annual and released with significant lag.
    - National accounts: GDP numbers are annual and released with a significant lag; quarterly GDP should be published and methodology made more transparent.
    - Timeliness for consumption, saving, employment, and inflation should be less than one month after reference period.
    - Agricultural statistics are released with almost one year’s lag.
  - Accessibility, metadata, and revision policy
    - Most publications do not carry methodological descriptions; explanatory notes are often inadequate or vague.
    - Absence of an advanced release calendar.
    - Revision policy for production and national accounts data is not disseminated; revisions to GDP and government finance often unexplained.
    - Difficult to get data on request; no concept of accepting e-mail requests in some government bodies.
    - Government websites not always easily or swiftly accessible or regularly updated; not all datasets downloadable.
    - Recommendation: provide methodological notes with data and make all datasets regularly updated and downloadable.
  - Overall assessment and institutional recommendations
    - Financial and debt statistics from the SBP are accurate, comprehensive, and timely; MOF data credibility is of concern.
    - Official statistics contain many revisions; final figures are announced too late.
    - Official statistics sometimes inconsistent across agencies (e.g., FBS trade data differ from SBP or Pakistan Customs; FBS sources do not follow harmonized system codes).
    - Calls for improved coordination among agencies to eliminate duplication.
    - Suggestions to involve academic institutions and researchers in data collection and dataset design.
    - Recommendations to disseminate an advanced release calendar via print media.
    - Proposals to establish independent agencies to ensure unbiased data and for the Federal Bureau of Statistics to undertake serious reforms (survey methodology, scope, frequency).
    - Specific topical needs identified: more frequent labor force surveys (currently every three years), improved GDP sector estimates (small scale manufacturing and livestock), periodic surveys for informal economy, better poverty data, and updates to national accounts methodology and base year.

- Selected quantitative survey results (selected items as provided)
  - Question: Which official statistics do you use regularly? (examples of counts shown in table; full table preserved in source)
  - Question: Do you refer to official descriptions of the sources and methods? — Yes 44 No 43 No opinion 0 (counts by category available in table).
  - Question: Are you satisfied with the coverage of official statistics? — Overall Yes 44 No 43 No opinion 0 (category breakdowns in table).
  - Question: Are you satisfied with the timeliness of official statistics? — (category responses summarized in table; see source for full counts).
  - Question: Do you know if there is a publicly disseminated calendar that announces in advance the dissemination dates for official statistics? — Responses summarized in table.
  - Question: How do you access the overall quality of the official statistics (1 rated as poor and 5 excellent)? — Overall weighted average rating 3.16; category ratings 3.29 3.50 3.60 2.00 3.57 2.60 3.00.

*Source: _cr04398 - 0.1 Legal and institutional*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04398.pdf_