0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

The 1994 Statistics Policy Declaration took stock of Burkina Faso’s national statistics system revealing the need to reorganize it, starting with the restructuring of the National Institute of Statistics and Demography (INSD) as the central statistical information body. Subsequently, a series of laws was adopted, providing for (1) the creation of the National Statistical Coordination Council (CNCS) and its composition; (2) the status of the INSD and its powers; and (3) the statistical secrecy and reporting obligations. The Decree No. 2000–508 of October 27, 2000 gave the INSD the status of a public administrative institution with legal personality and financial independence. The INSD reports to the Ministry of Economy and Development (MED) on technical matters and to the Ministry of Finance and Budget (MFB) on financial matters.

By virtue of that decree, the INSD is responsible for collecting, processing, and publishing the national accounts. In practice, however, the Directorate of Forecasting and Macroeconomic Analysis (DPAM) of the MED produces the initial estimates of Gross Domestic Product (GDP) in consultation with the INSD and other data producing agencies. These estimates are validated by the Committee for Forecasting and Multilateral Surveillance (CPSM), composed of DPAM, INSD, BCEAO, and the other data producing agencies.

0.1.2 Data sharing and coordination among data producing agencies are adequate

The national management and coordination body is the CNCS, with the INSD serving as its secretariat. Created in 1997 (Decree No. 97-371/PRES/PM/MEF), the CNCS is supposed to hold regular meetings, as well as special sessions when needed, and to adopt an annual program of statistical work. However, the Council has not functioned well because of lack of resources and technical support. Nor does INSD have sufficient resources to perform its secretarial tasks.

In addition to the financial and human constraints the INSD is facing, the lack of coordination at the national level and the shortcomings of the statistical apparatus, which is supposed to be decentralized, pose additional difficulties in the performance of its functions.

In 2003, the Director General of the INSD proposed a new organization chart to meet current needs. This proposal, which calls for restructuring the internal organization and strengthening the division responsible for cooperation, has not yet been adopted. Nevertheless, the INSD makes efforts to maintain good contacts with data providers. Cooperation with ministries and other data producing agencies may be considered as satisfactory. The significant delays in publishing data are rather related to the lack of financial and human resources at the INSD, as well as to the delays encountered by the data providers, primarily the ministries.

Workshops have been conducted on a fairly regular basis by the INSD itself, or by other ministries, where INSD is invited to participate. These meetings facilitate the exchange of information among the different national institutions. One such workshop that was held recently dealt with harmonizing the terminology for classifying economic activities.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

The Law No. 40/96/ADP of November 8, 1996, on reporting obligations and statistical secrecy, contains stipulations to protect the confidentiality of statistical information to maintain confidence in official statistics, and to guarantee respect of methodological and ethical rules. The law states that “individual economic or financial information collected may in no case be used for purposes of tax control or economic repression,” and that “information relating to individuals or corporations, with the exception of data on health and sexual life (…) may be transferred to the National Institute of Statistics and Demography or to the ministerial statistics offices and similar public institutions solely for the purpose of statistical compilation.” Although the authorities interpret this law as protecting the data provided by businesses, it is vague, particularly as regards certain categories of data.

In practice, the INSD performs special aggregations in order to protect the confidentiality of individual data. For example, the units belonging to the “electricity, gas, and water” sector are aggregated, as are those in the “transport, storage, and communication” sector.

The questionnaires that are sent out include references to Law No. 40/96/ADP, reminding respondents of their reporting obligations, and giving assurance that the information they provide will be used exclusively for statistical purposes. This law also stipulates that public officials “are subject to statistical secrecy under penalty of the sanctions provided by the criminal code relating to insult, slander, and disclosure of secrets. In this context, personnel engaged in statistics work must be sworn to secrecy.”

Access to each reporting entity’s data is limited to officials who need those data in the exercise of their duties. Only personnel of the Section of Economic Accounts (SCE) have access to the computers, and their offices are locked when they are absent.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

Pursuant to Law No. 40/96/ADP, “any individual or corporation must respond accurately and within the established time frames to official statistical surveys,” and they are liable to penalties for noncompliance. In practice, however, these prerogatives are not usually applied in the case of nonresponse, with the INSD preferring to secure the voluntary cooperation of respondents.

Much of the statistical information is collected by other government agencies that are part of the decentralized statistics system in Burkina Faso. However, as the central statistical information body, the INSD is mandated to request the information collected by any other statistics-producing bodies and data providers, notably government agencies.

Out of consideration for respondents’ workload, the SCE personnel assist them in completing and submitting the questionnaires. Moreover, the INSD tries to maintain good cooperation with respondents, and uses incentives such as providing them with data they may request, or occasionally sending them a copy of the publication free of charge (particularly in the case of ministries and other government agencies).

0.2.1 Staff, financial, and computing resources are commensurate with statistical programs

Currently, the SCE staff consists of nine persons, of which two economist-statisticians (one of whom is the head of the section), three statistical analysts, and four technical assistants. The section is part of the Directorate of Economic Studies, whose Director coordinates the SCE’s work. The staff members are licensees of the statistics schools of Abidjan, Kigali, or Niamey. The current staff number needs to be increased in light of the work required to accomplish the project of modernizing the national accounts in due time.

The DPAM staff includes five-six agents. This staff, responsible for the regular production (quarterly) of the trend chart for the main macroeconomic aggregates, as well as short- and medium-term forecasts, seems adequate for its tasks.

In terms of staff training, every staff member of the SCE has participated in seminars or short courses on national accounts methodology, conducted by regional agencies. All staff members have participated in the training within the project for the implementation of the ERE/TES module (supply-use/input-output tables). Often, the positions in the SCE are occupied by staff who have previously worked in other sections of the INSD. Promotions are generally based on merit. However, recent financial constraints within the INSD have led to a high turnover, adversely affecting the available professional skills.

The incentive provisions are those of the public service. Salaries and bonuses seem very low in light of the need to retain qualified personnel. The INSD has lost much of its staff in the last few years, particularly senior personnel. For example, the current staff of the INSD counts about 45 officers as compared to over 100 in 1995.

The available computing equipment has been provided through external assistance; the SCE now has eight computers of good capacity. The computer maintenance budget, however, is too low, as is the overall operating budget of the INSD.

0.2.2 Measures to ensure efficient use of resources are implemented

The annual operating budget of the INSD, which is approved by its Board of Directors, is distributed by action, taking into account the budgetary amount allocated and the prospects of securing supplementary funding from external assistance programs. As part of the coordination of the INSD activities, strategic plans are shared with the staff.

The concepts and methodologies used are consistent with those of national accounting. In the course of conducting surveys, all the relevant elements (methodology, questionnaires, standards, etc.) are discussed in order to come up with concrete proposals for requesting external assistance. Its financial dependence on external support, however, leads the INSD to giving priority to donor-driven activities, at the expense of important regular activities such as statistical surveys.

The INSD monitors the cost of statistical surveys and censuses, and measures are foreseen to ensure that available resources (heavily dependent on the support of donors and lenders) are being used efficiently.

0.3.1 Processes are in place to focus on quality

Data quality is a major concern of the INSD, but it finds itself constrained by a number of factors that affect the production of good statistics within reasonable time frames. There are no trade-offs between the various dimensions of quality. External audits of the methods employed are conducted in the context of technical and financial assistance projects.

The INSD program for the improvement of its statistics includes as the primary objective the modernization of the national accounts, with the implementation of the ERE/TES module.

0.3.2 Processes are in place to monitor the quality of the collection, processing, and dissemination of statistics

Quality control over the collection and processing of data is closely monitored by the SCE staff.

The official body responsible for coordinating statistical activities in Burkina Faso, the CNCS, has not worked well, and its last meeting dates back to December 2001. The purpose of that meeting was to analyze results from the census of existing committees and commissions in the country and their operational status, with a view to reactivating the specialized technical apparatus for coordinating national statistics collection and processing activities. The will to do so, however, is not consistently evident. Meanwhile, the INSD has taken on the responsibility of approving statistical surveys and censuses relating to public services.

A mechanism to monitor user reactions and identify emerging needs has not been implemented. The introductory notes in the publication “National Economic Accounts” invite users to express their opinions and suggestions on the contents of publications, but responses are not closely monitored.

The managers of the INSD recognize the importance of this aspect and are planning to formalize producer-user relations.

0.3.3 Processes are in place to deal with quality considerations, including trade-offs within quality, and to guide planning for existing and emerging needs

A number of factors limit the trade-offs among the various dimensions of data quality at the INSD. Certain aspects of data quality, in particular those relating to data reliability, are followed with priority to the detriment of their accessibility and serviceability. The failure of other statistics producing agencies and data suppliers to provide the information within acceptable time frames is another factor delaying the production and publication of statistics by the INSD.

No user survey has been organized to get feedback on data quality issues. Meetings with users are organized only on the occasion of population censuses and large-scale surveys, in order to inform the public about the objectives of these actions, and to enlist their cooperation.

1.1.1 Statistics are compiled on an impartial basis

The INSD is administered by a council with representatives of the main data-producing agencies. While the 2000 Decree conferring the INSD the status of a public administrative institution stipulates that its Director General has very broad powers to act on behalf of the Board of Directors, and to hire and discharge staff, the decree contains no reference to the INSD’s professional independence, or to prohibit outside interference regarding the content or disclosure of statistical data. Nevertheless, the management of the INSD feels that it is able to operate with complete professional independence.

The INSD staff is encouraged to attend seminars and meetings with other professional groups, and to publish methodological papers.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

Statisticians are free to choose the most relevant data sources and methodology, but they are constrained by resource considerations and depend on the financial support of regional or international donors and agencies. In fact, within the regional harmonization process of statistical methods for national accounts, all member countries of the West African Economic and Monetary Union (WAEMU) follow a common methodology.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The Director General of the INSD has the right to issue comments on erroneous interpretations or the misuse of statistics, but no such situation has been encountered to date.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The information regarding the conditions governing the collection, processing and dissemination of data can be found in the relevant legislation, which is publicly available (including on the Internet). Statistical publications identify the principal sources of data and the producing agencies.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

Government officials have access to data before their publication, but only for internal use in the preparation of economic and social policies.

This practice has not been made public and is not mentioned in the General Data Dissemination System (GDDS) metadata.

1.2.3 Products of statistical agencies/units are clearly identified as such

The INSD publications (National Economic Accounts, Statistical Yearbook, etc.) are identified as such. They are published under the name of the National Institute of Statistics and Demography, indicating the responsible ministry as well. Whenever the INSD statistics are used or reproduced by other agencies—primarily the Central Bank of West African States (BCEAO), the West African Economic and Monetary Union (WAEMU) and the Sub-Saharan African Economic and Statistical Observatory (AFRISTAT)—the source is always cited.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

Major changes in methodology or statistical techniques are not notified in advance. Currently, the INSD is in the process of changing the methodology as well as the base year for the national accounts. The public has not been informed of this work. Nevertheless, it should be noted that this work is being pursued in the context of projects with external support, which have been approved by the Council of Ministers, whose decisions are published in the Official Gazette.

1.3.1 Guidelines for staff behavior are in place and are well known to the staff

The rules of behavior are the same for all government employees, and are laid down in Law 010/98/98/AN of April 21, 1998, on comprehensive reform of the public administration.

In principle, the INSD personnel should take an oath to respect the provisions of the statistical secrecy law. However, this rule is not always observed.

Ethical standards are not routinely explained to new employees upon their arrival nor are personnel regularly reminded of them.

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

The general framework for compiling the national accounts is the System of National Accounts 1968 (1968 SNA). Work is underway to implement the System of National Accounts 1993 (1993 SNA), following a common methodology adopted by WAEMU for its member countries. The first accounts based on this new system, for 1999, should be completed shortly using the ERE/TES module.

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices

Because of constraints encountered in the production of statistical indicators, the national accounts are not available on an annual basis, although they are prepared on the basis of annual indicators. The latest accounts produced are those for 1997. Work is underway on the accounts for 1999–2001. This work is being pursued in parallel on two basis: (1) the new accounts, using the common WAEMU methodology, and (2) the 1968 SNA based accounts. For those years for which the national accounts are available, the series of 1993 SNA tables and accounts identified as “minimum requirements” by the Inter-Secretariat Working Group on National Accounts (ISWGNA) are not entirely complete (the methodology used follows the 1968 SNA recommendations). Specifically, the GDP expenditures at constant prices (1985) are not available, and the domestic activities of certain regional institutions is not covered.

The level of detail of the published data is limited to 19 industries, grouped into three categories: the informal sector, the modern private sector, and the modern public sector. The other aggregates produced and published include the production and generation of income accounts by sector of activity, the institutional sector accounts, and the general table of transactions. Supply and use tables are not produced. The first input-output tables are now in preparation for 1999 using the ERE/TES module; 1999 will become the new base year.

The national accounts cover all resident economic units in Burkina Faso and territorial enclaves in the rest of the world (embassies, consulates). They exclude the National Agency of the BCEAO and the Air Afrique.

The production and assets boundaries follow broadly the 1968 SNA.

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices

The classifications used by the INSD are generally compatible with international standards, but they are restricted to classifying economic activities and household consumption. The Central Product Classification (CPC) and the Classification of the Functions of Government (COFOG) are not used. To classify economic activities, the INSD uses a common classification for WAEMU countries, the Activity Nomenclature of the Member States of AFRISTAT (NAEMA), derived directly from the International Standard Industrial Classification of All Economic Activities (ISIC, Rev.2). This classification has been adapted to national circumstances, including: (1) a distinction between agriculture and livestock activities; (2) a fusion of the subsectors “Manufacture of Wood Products” and “Manufacture of Metal Products;” (3) a distinction between road and rail transport; (4) the abandonment of the distinction between retail and wholesale trade, although the oil trade is separated from the rest of trade; (5) a distinction between market services and nonmarket services; and (6) the units belonging to the branches “electricity, gas, and water” are aggregated, as are those in the branches covered under “transport, storage, and communication,” in order to protect data confidentiality. To classify household consumption, the INSD uses the Classification of Individual Consumption by Purpose (COICOP).

The majority of the nonprofit institutions serving households are grouped together with general government. The general government sector distinguishes between domestically and externally financed units.

2.4.1 Market prices are used to value flows and stocks

The rules employed for recording flows and stocks are consistent with the 1968 SNA. Market output is valued at factory gate prices, with all taxes including value-added tax (VAT), which is not consistent with the internationally accepted standards. Imports are valued c.i.f. (cost, insurance, and freight), and exports are valued f.o.b. (free on board). In principle, foreign currency transactions are converted at the average annual market exchange rate.

2.4.2 Recording is done on an accrual basis

Most transactions are recorded on an accrual basis, consistent with 1993 SNA. Central government receipts and disbursements, however, are recorded on a cash basis.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices

Recording is done on a gross basis. There are no estimates of fixed capital consumption. Transactions between establishments of the same enterprise are recorded on a gross basis, when the necessary information is available.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

A List of Industrial and Commercial Enterprises (RIC) was prepared in 1998. It covers all large firms submitting a Statistical and Fiscal Declaration (DSF) (according to a threshold that takes into consideration the volume of business) as well as small enterprises that do not provide DSFs. The RIC is updated only when a new census is conducted, which is not done annually (the next one is planned before the end of 2003).

Data sources are essentially administrative (DSF and other administrative data provided by ministries and specialized departments), supplemented by surveys (for example, the Survey 1–2–3, which includes sections for the informal sector, employment and household consumption but whose coverage is limited to the capital city; the priority survey on household consumption of national coverage; light surveys on the informal sector, etc.), which, because of funding constraints, do not follow a regular calendar. These surveys depend largely on technical and financial support from regional or international agencies.

The population census is conducted roughly every ten years. The last censuses took place in 1985 and 1996, and served as basis for the priority survey on household consumption. This survey is now underway, and will serve several purposes, including providing an estimate of final household consumption (which is currently derived residually) for the new national accounts series.

The coverage of the priority survey on household consumption is complete and representative for all geographic areas and all socio-income categories. The survey uses random sampling techniques consistent with international standards, and the techniques used for calculating indicators are appropriate. The data collected are sufficiently detailed in terms of acquisitions, investments, employment, etc.

Since surveys are not conducted annually, the INSD uses the quarterly DSFs to compile the accounts for the modern sector units. The DSF response rate is satisfactory (between 70 and 85 percent), but there may be delays in their collection. The INSD should identify appropriate information sources for preparing provisional accounts. The quarterly survey for the industrial production index could be an important source, in addition to the other quarterly information collected.

The following table shows the principal data sources for the national accounts by group of industries, and their weight in GDP.

Production Approach

Production Approach

Activity	Weight in GDP	Principal data sources
Agriculture, forestry, and fishing	35%	Crops (for own use and market): - Agricultural Statistics Division of the Ministry of Agriculture (farm census, Standing Agricultural Survey, which provides annual data on area under cultivation and on yields); -National Food Security Stocks Management Company (SONAGESS); -Textile Fiber Company–SOFITEX, for cotton production; -Cereal prices from the Market Information System (SIM) and price surveys for the CPI;
		Livestock: -Ministry of Animal Resources; -Price statistics provided by the INSD Price Section;
		Forestry, fishing, hunting: -Estimates based on administrative statistics supplied by the specialized units of the Ministry of Environment and Water; -Imports provided by the Market Output of Trade and Services Section of the INSD.
Mining and manufacturing	14%	- DSFs; -Indirect estimates for modern non-DSF units; Informal sector: estimates based on the informal sector survey and the consumption budget survey.
Electricity, gas, and water	1%	Accounting records of companies belonging to the industry.
Construction and public works	5%	Public works: DSFs for companies; Construction: DSFs supplemented with estimates for modern non-DSF units Informal sector: extrapolations based on the informal sector survey
Wholesale and retail trade	15%	-DSFs from companies in the modern sector; -Estimates for modern non-DSF units; -Various price statistics prepared by INSD price section; -Light survey of trading margins
Transport and communication	4%	Rail transport: accounting records of companies in the business Road transport: DSFs and extrapolations based on the informal sector survey; Air transport: accounting records of the national airline company Allied services: DSFs and estimates for modern non-DSF units; Communication: accounting records of companies in the business
Financial institutions & insurance companies	2%	-DSFs of companies in the business
Other market services	10%	-DSFs; -Treasury data on operating accounts; -Extrapolations based on the informal sector survey
General government	11%	Domestically financed: data from the Ministry of Finance (statements of receipt, remittance slips, budgetary status of the General Treasury, Government Operations Table (TOFE), provincial operating accounts, taxes, and fees); -Activity reports from the largest companies Externally financed: United Nations Development Program (UNDP) annual report; -Balance of payments, -Other project monitoring institutions (Nongovernmental Organizations—NGOs)
Nonprofit institutions serving households	-	The great majority of these units are treated as externally financed general government.
Import taxes and duties net of subsidies	3%	-Documents from the Ministry of Finance; -Customs department.
Expenditure Approach
Final household consumption	77%	Residual
Government final consumption	14%	Domestically financed: Data from the Ministry of Finance; Externally financed: estimates based on the UNDP annual report, the balance of payments, etc.
Gross-fixed-capital formation	20%	DSFs of companies in the modern sector; Public investment program; Accounting records of local government and NGOs
Changes in stocks	~1%	DSFs of corporations and quasi-corporations
Imports and exports of goods and services	-12%	Customs statistics; Balance of payments

View Table

Production Approach

Activity	Weight in GDP	Principal data sources
Agriculture, forestry, and fishing	35%	Crops (for own use and market): - Agricultural Statistics Division of the Ministry of Agriculture (farm census, Standing Agricultural Survey, which provides annual data on area under cultivation and on yields); -National Food Security Stocks Management Company (SONAGESS); -Textile Fiber Company–SOFITEX, for cotton production; -Cereal prices from the Market Information System (SIM) and price surveys for the CPI;
		Livestock: -Ministry of Animal Resources; -Price statistics provided by the INSD Price Section;
		Forestry, fishing, hunting: -Estimates based on administrative statistics supplied by the specialized units of the Ministry of Environment and Water; -Imports provided by the Market Output of Trade and Services Section of the INSD.
Mining and manufacturing	14%	- DSFs; -Indirect estimates for modern non-DSF units; Informal sector: estimates based on the informal sector survey and the consumption budget survey.
Electricity, gas, and water	1%	Accounting records of companies belonging to the industry.
Construction and public works	5%	Public works: DSFs for companies; Construction: DSFs supplemented with estimates for modern non-DSF units Informal sector: extrapolations based on the informal sector survey
Wholesale and retail trade	15%	-DSFs from companies in the modern sector; -Estimates for modern non-DSF units; -Various price statistics prepared by INSD price section; -Light survey of trading margins
Transport and communication	4%	Rail transport: accounting records of companies in the business Road transport: DSFs and extrapolations based on the informal sector survey; Air transport: accounting records of the national airline company Allied services: DSFs and estimates for modern non-DSF units; Communication: accounting records of companies in the business
Financial institutions & insurance companies	2%	-DSFs of companies in the business
Other market services	10%	-DSFs; -Treasury data on operating accounts; -Extrapolations based on the informal sector survey
General government	11%	Domestically financed: data from the Ministry of Finance (statements of receipt, remittance slips, budgetary status of the General Treasury, Government Operations Table (TOFE), provincial operating accounts, taxes, and fees); -Activity reports from the largest companies Externally financed: United Nations Development Program (UNDP) annual report; -Balance of payments, -Other project monitoring institutions (Nongovernmental Organizations—NGOs)
Nonprofit institutions serving households	-	The great majority of these units are treated as externally financed general government.
Import taxes and duties net of subsidies	3%	-Documents from the Ministry of Finance; -Customs department.
Expenditure Approach
Final household consumption	77%	Residual
Government final consumption	14%	Domestically financed: Data from the Ministry of Finance; Externally financed: estimates based on the UNDP annual report, the balance of payments, etc.
Gross-fixed-capital formation	20%	DSFs of companies in the modern sector; Public investment program; Accounting records of local government and NGOs
Changes in stocks	~1%	DSFs of corporations and quasi-corporations
Imports and exports of goods and services	-12%	Customs statistics; Balance of payments

Domestically financed general government accounts include central government (general budget, special treasury accounts, and nonbudgetary transactions), miscellaneous government agencies (stabilization funds, offices, projects, Chamber of Commerce, and other organizations), local governments (provincial and municipal), and social security funds.

Data sources for government accounts are the statements of receipt and remittance slips for the national budget, as well as the budgetary status of the General Treasury. Certain transactions pending allocation, classification, or settlement are analyzed as nonfinancial transactions. Nonbudgetary transactions are identified by comparing overall receipts and expenses as they appear in the budgetary status of the General Treasury against the statements of receipt and remittance slips.

Externally financed general government accounts include those of nonprofit institutions serving households, which are themselves essentially financed from abroad. The principal sources of information are the United Nations Development Program (UNDP) annual report on cooperation and development, the balance of payments, as well as the unit responsible for monitoring projects and those responsible for monitoring the activities of nongovernmental organizations.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

For the informal sector (which accounts for 35 to 40 percent of national economic output, without counting agriculture), there is little current information available. Due to financial constraints, only light surveys of the informal sector are conducted on an irregular basis (the most recent date from 1989 and 1997). Data collected through these surveys on sectors of activity are used to compile the production account for the current year and as basis of extrapolation for the following years. Unfortunately, due to the limited coverage of the last survey to the capital city (about 8 percent of the country’s population), it can only be used with caution.

Generally, source data for the modern sector are consistent with national accounting needs. When it comes to agriculture, the crop year runs from October to the following September, thus straddling two years (n and n+1). The first portion of the crop year (October-December), representing the three harvesting months, is allocated one-third of the output of the crop year n/n+1. The remaining two-thirds are allocated to the output of the calendar year n+1. This practice is consistent with the 1993 SNA requirements.

Most transactions are recorded on an accrual basis. However, central government receipts and disbursements are recorded on a cash basis.

The BCEAO makes adjustments to foreign trade figures in the balance of payments, to cover the unrecorded border trade. The adjusted figures are subsequently brought into the national accounts.

3.1.3 Source data are timely

The delay in data collection is one of the main factors behind the delays experienced in preparing and publishing the national accounts: the last available series of final accounts are those for 1989–93, and the last semi-final accounts are for 1994–97.

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

Given the weak data sources for the informal sector, there is no reliable and generally accepted estimate of the scale of unrecorded activities in the national accounts. Certain corrections are made to foreign trade data provided by customs to reclassify activities and to cross-check their data on the basis of the information from certain enterprises. INSD makes no adjustment for the unrecorded border trade.

3.3.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide planning

Most data are provided from administrative records (the DSFs and the other administrative records provided by directorates and other administrative agencies), supplemented by occasional surveys. For these surveys, the INSD has information available on sampling errors, which are examined along with the other survey parameters.

For administrative data, data from the field are analyzed and any errors detected are corrected by contacting the concerned companies, or during the collection process. Data are also compared against information from previous years for the same companies. Any aberrations are identified and corrected, and any suspect values are confirmed with the reporting company. The accuracy of budgetary data is also validated regularly, and explanations are requested in case of doubt. However, in the case of foreign trade statistics, there is no adjustment to unrecorded border trade. Certain adjustments are made by the BCEAO, which are taken as such in the national accounts.

There is no regular analysis of the coverage of basic data sources.

3.4.1 Main intermediate data are validated against other information where applicable

There is little secondary information available for validating data. GDP estimates are determined from both the production and the expenditure sides. Given the preponderance of production-approach estimates, there are few discrepancies among the various national accounts aggregates (final household consumption being a residual). An adequate framework, such as supply-use tables, is not yet in place to validate the internal consistency of data. These tables are to be produced using the ERE/TES module, beginning with 1999.

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated

Verifications are conducted to detect any discrepancies in intermediate data, particularly for administrative data. Ratios of intermediate consumption to production, structure of GDP, etc. are analyzed and discrepancies are eliminated.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

Production and generation of income accounts are prepared by economic sector of activity (group of industries), but no supply-use tables are established. Because GDP is determined primarily from the production approach, there are no discrepancies between GDP by type of activity and GDP by category of expenditure.

3.5.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

Under normal circumstances, the INSD produces four versions of the national accounts for any given year n: a provisional version (by the end of the following year), a semi-final I version (by the end of year n + 2), a semi-final II version (by the end of year n + 3), and the final version (by the end of the year n + 4). The current situation is rather special, because there is a series of final accounts up to 1993, and a series of semi-final II accounts up to 1997. Efforts are now focused on preparing the accounts for 1999 using the ERE/TES module, and consequently work on the other intermediate year accounts has been virtually suspended.

With respect to revisions between successive versions for a given year, it must be noted that the various versions of accounts for a given year are based on the same methodology and use nearly the same data sources. Early estimates are not produced for the provisional accounts, as should normally be done. Because of the concern that it is impossible to change the figures once the accounts are published in their final version, the INSD publishes a semi-final II version of accounts, delaying therefore the dissemination of the final version. For example, the last series of accounts for 1994–1997, which should have been regarded as final, came out in a semi-final version. There is no study or analysis of revisions between successive versions of data.

On the other hand, the results of the revisions of DPAM’s provisional estimates are well documented. However, there is no investigation of any discrepancies occurring between the provisional estimates using the IAP and the accounts prepared by the INSD.

4.1.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored.

There is no regular mechanism for consultation with ministries. Meetings are limited to occasional sessions for coordinating methodological changes or harmonizing standards and codes where the involvement of other ministries and central agencies is required.

There are no systematic procedures in place for consulting the general public. At the suggestion of users, the national accounts published for the years 1985–1992 were accompanied by supplementary tables and methodological notes for constant-price estimates. This practice has not been followed since, however.

4.2.1 Timeliness follows dissemination standards

There are significant delays in the publication of the national accounts. The semi-final II version for the years 1994–1997 was published only in March 2001. The number of copies is very limited, and is mainly intended for major institutional users. The general public can only consult these documents at the INSD. On the other hand, the WAEMU publishes a regular series of principal aggregates for its member countries, but the timeliness of the data depends on their production by the INSD.

The INSD is planning to finalize and publish the 1999 accounts on the new basis shortly. A methodology for estimating early provisional accounts should be introduced as promptly as possible.

To meet the needs for up-to-date data on the part of decision-making bodies and other users, the DPAM of the MED produces, in cooperation with the INSD and other data producing agencies, the initial GDP estimates.¹ DPAM disseminates these and updated estimates regularly (every three or four months), but selectively.

4.2.2 Periodicity follows dissemination standards

The national accounts are prepared on the basis of annual indicators. Because of constraints in the production of statistical indicators, however, they are not available every year.

4.3.1 Statistics are consistent within the dataset

GDP is estimated primarily from the production approach. There is no independent assessment of GDP from the expenditure approach, and final household consumption is calculated as a global residual in the final version of the accounts. The semi-final accounts for the most recent years contained certain discrepancies among the indicators produced, but these will be eliminated in the final version.

GDP estimates are provided at current prices and at constant (1985) prices, consistent with the formula value = volume x price.

4.3.2 Statistics are consistent or reconcilable over a reasonable period of time

The INSD disposes of consistent time series for the period 1985–1997 (with 1985 as the base year). There has been no change of methodology or of base year for this series. A new series of accounts, having 1999 as base year, will be prepared for the years 1999 onwards, leaving 1998 as provisional date. This will be done as part of the WAEMU regional project for adopting the 1993 SNA methodology.

The INSD is planning to calculate retrospective series, but this work has been given secondary priority in light of other demands.

The published national accounts do not contain any analysis of the trend of economic indicators, and there is no commentary on adjustments made to maintain the consistency of data over time.

4.3.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

The statistics in the national accounts are largely consistent with those for the balance of payments and for government finance since they are largely based on the same sources. Estimates of imports/exports of goods are corrected by the INSD (using the EUROTRACE module) and are sent to the BCEAO for final reconciliation, and to ensure the consistency of the indicators. The BCEAO makes other specific corrections, as well as corrections to improve statistical coverage by taking into account the impact of unrecorded border trade.

In terms of consistency with government finance statistics, the INSD verifies the consistency of the figures and adjusts them as needed to bring them into line with indicators in the Government Operations Table (TOFE).

Consistency between the different data series (GDP estimates, balance of payments, TOFE, etc.) produced by DPAM is not fully guaranteed. Data series are regularly adjusted on the basis of more recent information.

4.4.1 Revisions follow a regular, well-established, and transparent schedule

Although there is a pre-established cycle for revising the series of accounts produced by INSD, it has not been made public and it is not in fact respected (see also 3.5.1). Because of the lengthy delays in producing and publishing the data, economic policy managers often resort to estimates made by the DPAM using the IAP module. These data are revised on a quarterly basis, and published after their validation by the CPSM.

4.4.2 Preliminary data are clearly identified

Preliminary versions of data are clearly identified as such whenever they appear in the INSD publications (for example, in the national economic accounts for 1985-1992, published in November 1993). They are published under the authority of the INSD and the MED.

4.4.3 Studies and analyses of revisions are made public

Changes resulting from revisions are examined but the methodological reasons explaining them are not made public. This information is available on request at the INSD.

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

Estimates of the national accounts are published at a reasonable level of detail, including chronological series, in a distinct publication, the “National Economic Accounts.” The information is adequately presented in tables that can be easily read. Moreover, the statistical yearbook contains the principal aggregates of GDP (prepared by INSD) as well as the expenditure components of GDP compiled by the DPAM of MED. Nevertheless, there is room for improvement in the published statistics, particularly by including analyses of the trend of certain data categories from one period to another.

Supplementary tables and methodological notes for estimates at constant prices have been introduced at the suggestion of users.

In preparation for data validation meetings, the DPAM produces a booklet entitled “Economic Situation of Burkina Faso in (year) and Outlook for (period),” which contains some analysis of important indicators.

5.1.2 Dissemination media and formats are adequate

Because of financial constraints, the means of dissemination fall short of users’ needs. The data appear at long and irregular intervals. Publication is not accompanied by any press release. The INSD does not have a web site, but the main aggregates are posted on the WAEMU site. The INSD is planning to set up its own web site with financial assistance from France and technical support from AFRISTAT.

The public may consult older chronological series upon request.

5.1.3 Statistics are released on a preannounced schedule

Statistical series are published only irregularly. Due to the lack of resources, the release dates for the next publications, even if announced in certain publications, are not respected.

5.1.4 Statistics are made available to all users at the same time

The government and the ministries responsible for collecting and producing statistics have access to the data before publication, but only for internal use and economic policy purposes.

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request

Data are not widely disseminated among the public at large. Certain categories of users have priority of access to more detailed data. However, data for industries in which there is only one enterprise are not provided, for reasons of statistical confidentiality.

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

The metadata are of only limited accessibility. Methodological notes on concepts, classification, sources and methods for preparing the national accounts are brought to the attention of users via the publication “National Economic Accounts.” However, because of financial constraints, the print run of this publication is inadequate, although the information may be consulted in the INSD documentation room.

5.2.2 Levels of detail are adapted to the needs of the intended audience

There is no comprehensive document available on the sources and methods used to compile national accounts. Some methodological notes are included in the publication of the economic accounts for 1989–1993), but these are not updated. The GDDS metadata are posted only on the IMF web site, in English, a fact that is not even brought to the public’s attention.

5.3.1 Contact person for each subject field is publicized

The INSD has a public sales office, but its publications do not identify the contact person for further information.

5.3.2 Catalogs of publications, documents, and other services, including information on any charges, are widely available

A list of publications, indicating their price, is available at the INSD sales office. Assistance to users in placing orders is available.

Table 1.

Burking Faso: Data Quality Assessment Framework—Summary of Results for National Accounts

(Compiling Agency: National Institute of Statistics and Demography, INSD)

Table 1.

Burking Faso: Data Quality Assessment Framework—Summary of Results for National Accounts

(Compiling Agency: National Institute of Statistics and Demography, INSD)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			In practice, the initial estimates of GDP are produced by DPAM, under the supervision of the CPSM, and not by INSD, contrary to the latter’s mandate. Weak coordination at the national level
	0.2 Resources				X		Human and financial resources are insufficient. Turnover of personnel is high.
	0.3 Quality awareness			X			No systematic procedure for quality improvement. Priority is given to accuracy and reliability of data to the detriment of their serviceability and accessibility.
1. Integrity
	1.1 Professionalism			X			INSD’s professional independence is not enshrined in law.
	1.2 Transparency			X			Access of government officials to data before publication is not made public.
	1.3 Ethical standards			X			Ethical standards not routinely made known.
2. Methodological soundness
	2.1 Concepts and definitions			X			1968 SNA is applied	Work is underway to adopt the 1993 SNA. The first accounts based on the new system, for 1999, should be completed shortly, using the ERE/TES module.
	2.2 Scope				X		GDP by expenditure at constant prices is not compiled. Production and asset boundaries are too narrow. The scope excludes some regional units.
	2.3 Classification/sectorization				X		Sectorization is not consistent with 1968 SNA. The central product classification (CPC) and COFOG are not implemented.
	2.4 Basis for recording			X			Valuation of market output is inadequate. Government transactions are recorded on a cash basis.
3. Accuracy and reliability
	3.1 Source data				X		No regular program of surveys. Surveys of the informal sector are irregular and of limited coverage.	A priority survey of household consumption is now underway. Production of a harmonized index of industrial production is planned for the medium term.
	3.2 Statistical techniques				X		Level of compilation is too aggregated. GDP is calculated only from the production approach. Obsolete base year and weak constant price estimates.	The change of base year from 1985 to 1999 is planned with adoption of 1993 SNA.
	3.3 Assessment and validation of source data			X			Data source coverage is not routinely assessed.
	3.4 Assessment and validation of intermediate data and statistical outputs			X			Limited independent information to support validation procedures. Supply and use tables are not used to address discrepancies in data.
	3.5 Revision studies				X		Data sets are not produced on a regular basis. As a result, revision studies between various versions of data are not carried out consistently.
4. Serviceability
	4.1 Relevance			X			No formal consultations with users.
	4.2 Timeliness and periodicity				X		Various versions of national accounts are disseminated with big lags (over 4 years for the first data set) by INSD. Provisional GDP estimates are released regularly by the DPAM.
	4.3 Consistency			X			Consistency among certain aggregates is not strictly pursued.
	4.4 Revision policy and practice				X		Revision schedule is not transparent and documented.
5. Accessibility
	5.1 Data accessibility				X		Dissemination means are inadequate. Large deviations from expected publication dates. Published data are not accompanied by analyses.	INSD plans to set up a web site with external financial and technical support.
	5.2 Metadata accessibility			X			Metadata are not available to a wide public.
	5.3 Assistance to users			X			Except for GDDS metadata, published documents do not inform on contact person.

View Table

Table 1.

Burking Faso: Data Quality Assessment Framework—Summary of Results for National Accounts

(Compiling Agency: National Institute of Statistics and Demography, INSD)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			In practice, the initial estimates of GDP are produced by DPAM, under the supervision of the CPSM, and not by INSD, contrary to the latter’s mandate. Weak coordination at the national level
	0.2 Resources				X		Human and financial resources are insufficient. Turnover of personnel is high.
	0.3 Quality awareness			X			No systematic procedure for quality improvement. Priority is given to accuracy and reliability of data to the detriment of their serviceability and accessibility.
1. Integrity
	1.1 Professionalism			X			INSD’s professional independence is not enshrined in law.
	1.2 Transparency			X			Access of government officials to data before publication is not made public.
	1.3 Ethical standards			X			Ethical standards not routinely made known.
2. Methodological soundness
	2.1 Concepts and definitions			X			1968 SNA is applied	Work is underway to adopt the 1993 SNA. The first accounts based on the new system, for 1999, should be completed shortly, using the ERE/TES module.
	2.2 Scope				X		GDP by expenditure at constant prices is not compiled. Production and asset boundaries are too narrow. The scope excludes some regional units.
	2.3 Classification/sectorization				X		Sectorization is not consistent with 1968 SNA. The central product classification (CPC) and COFOG are not implemented.
	2.4 Basis for recording			X			Valuation of market output is inadequate. Government transactions are recorded on a cash basis.
3. Accuracy and reliability
	3.1 Source data				X		No regular program of surveys. Surveys of the informal sector are irregular and of limited coverage.	A priority survey of household consumption is now underway. Production of a harmonized index of industrial production is planned for the medium term.
	3.2 Statistical techniques				X		Level of compilation is too aggregated. GDP is calculated only from the production approach. Obsolete base year and weak constant price estimates.	The change of base year from 1985 to 1999 is planned with adoption of 1993 SNA.
	3.3 Assessment and validation of source data			X			Data source coverage is not routinely assessed.
	3.4 Assessment and validation of intermediate data and statistical outputs			X			Limited independent information to support validation procedures. Supply and use tables are not used to address discrepancies in data.
	3.5 Revision studies				X		Data sets are not produced on a regular basis. As a result, revision studies between various versions of data are not carried out consistently.
4. Serviceability
	4.1 Relevance			X			No formal consultations with users.
	4.2 Timeliness and periodicity				X		Various versions of national accounts are disseminated with big lags (over 4 years for the first data set) by INSD. Provisional GDP estimates are released regularly by the DPAM.
	4.3 Consistency			X			Consistency among certain aggregates is not strictly pursued.
	4.4 Revision policy and practice				X		Revision schedule is not transparent and documented.
5. Accessibility
	5.1 Data accessibility				X		Dissemination means are inadequate. Large deviations from expected publication dates. Published data are not accompanied by analyses.	INSD plans to set up a web site with external financial and technical support.
	5.2 Metadata accessibility			X			Metadata are not available to a wide public.
	5.3 Assistance to users			X			Except for GDDS metadata, published documents do not inform on contact person.

0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

Decree 2000-508/PRES/PM/MEF of October 27, 2000 (JO No. 46 2000) setting up the National Institute of Statistics and Demography (INSD) as an administrative government body is the most recent regulation that defines the functions of the INSD. Although the Decree contains no indication that the INSD is responsible for the compilation, processing, and publication of a consumer price index, it is generally accepted that the INSD, as the official government body responsible for statistical data, assumes these duties. The INSD budget explicitly allocates amounts for the compilation of the consumer price index.

On the basis of data collected by the INSD as well as prices collected by its own enumerators, the Central Bank of West African States (BCEAO) also compiles a consumer price index. However, this latter index is used solely for the BCEAO’s internal needs and is not published. It appears that the BCEAO requires consumer price data a few days before the results of the official index are released and that it needs a number of specifications that are not published by the INSD. A third reason is that price data collected by the Central Bank could also replace the official data in cases of urgency. Nevertheless, the existence of two largely parallel processes for price collection and the compilation of a CPI for Burkina Faso raises questions about the effectiveness with which resources are used.

0.1.2 Data sharing and coordination among data producing agencies are adequate

Generally speaking, the body responsible for managing and coordinating the national statistical data system in Burkina Faso is the National Statistical Coordination Board (CNCS). The INSD serves as the permanent secretariat of this body. Because of a lack of resources and member mobility, the CNCS, which was supposed to draw up an annual statistics program, does not function as anticipated. The Board last met in December 2001. The statistical system consists primarily of a group of producers operating independently from each other, the weaknesses of this system were clearly described in a paper entitled La Stratégie Nationale en Matière d’Information Statistique [The National Statistics Strategy] (1994). The fact that the CNCS is not functioning has therefore had a serious impact on the coordination of statistics and especially on the national accounts, which are dependent on such coordination.

Meanwhile, the consequences for the CPI are minimal. The INSD compiles the CPI essentially on the basis of data that it collects itself. The main data received directly from other institutions are mail, water, and electricity rates, which are supplied routinely by producers. Since the market was liberalized, fuel prices have no longer been treated as rates and are now collected by INSD enumerators from the distributors of these products.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

Law 040/96/ADP of November 8, 1996 states that, subject to the provisions of the Criminal Code on the breaking of seals and the removal of documents from public depositaries, individual data in collected documents bearing the statistical certificate and related to personal and family life or generally to private facts and behaviors may not be the subject of any communication from the staff of depositaries for a period of 100 years following the conduct of the related survey. Furthermore, in no case may individual economic or financial data be used for tax audit or economic repression purposes.

The authorities interpret this law as protecting the individual data supplied by enterprises, especially those data that are assumed to be sensitive, such as sales figures. Data of other types, such as the prices or exports of a specific enterprise, are not necessarily treated as statistical secrets.

The statistics certificate, which is attached to all survey forms, informs respondents of their rights and reporting obligations and assures them that the information they provide will be used for statistical purposes only. However, the persons in charge at the points of sale where prices for the CPI are collected are not informed from time to time of their rights and obligations.

The INSD provides the basic price data to the BCEAO, the WAEMU, and Afristat. These actions are legally justifiable on the basis of Articles 12 and 13 of decree 040/96/ADP, which allows that public or private official professional bodies, whether national or foreign, be authorized to serve as intermediaries or contractual agents in the performance of statistical studies and surveys on condition that these agents observe statistical secrecy. As stated, the BCEAO uses the source data for compiling its consumer price index (in addition to the prices collected by the central bank itself). The WAEMU provides a protected storage facility for the data collected by the statistical services in member countries. AFRISTAT receives data within the framework of its technical assistance in this area.

Staff of government units and contractual bodies authorized to serve as intermediaries for statistical studies and surveys are required to observe statistical secrecy, subject to the sanctions described in the Criminal Code. Statistics staff are required to take an oath in this regard, but in practice this obligation is not always met.

The INSD staff lock their offices when leaving. The INSD offices are guarded at night, but there are no janitors or guards at the doors. Computers are not protected with passwords. But there is also no central network of data, which limits access to data by staff other than those needing them for the performance of their duties.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

Pursuant to Article 4 of Law 040/96/ADP on the obligation to respond and statistical secrecy in Burkina Faso, individuals and legal entities are required to answer accurately and within the specified time limits to official statistical surveys. Any statistical survey ordered by a government unit is deemed to be official. In cases of refusal, a fine may be imposed, in an amount ranging from CFAF 10,000 (about US$15) for a first violation by an individual and CFAF 5 million (about US$7,500) by a legal entity committing a repeat offense. Fines to which offenders are subject are determined by the President of the CNCS. Decree 97–371/PRES/PM/MEF makes the Arbitration Commission of the CNCS responsible for settling disputes resulting from the implementation of Law 040/96/ADP. In practice, the INSD does not seek to force individuals and legal entities to respond to questionnaires through legal action.

There are no arrangements in place for taking account of the burden of work imposed on respondents, and the very decentralized national statistics system is not conducive to the consideration of such concerns. In theory, monitoring of the burden of work imposed on respondents could be a task for the CNCS. The INSD does not make any particular effort to obtain the cooperation of respondents by making them aware of the importance of good-quality statistics. However, the authorities participate in the African Statistics Days, when held.

0.2.1 Staff, financial, and computing resources are commensurate with statistical programs

Generally speaking, the ability of the INSD, as the official body responsible for statistics, to perform its duties laid down in the law is seriously undermined by a sharp drop in the number of statistics staff. Since 1995 their number has declined from more than 100 to 45, partly because of low remuneration and partly as a result of difficult working conditions. In the absence of prompt and decisive measures, it is even to be feared that the INSD will no longer have a critical number of trained personnel.

The Prices Unit comprises five staff members (four of whom serve mainly as enumerators), which is deemed insufficient. Especially in the two weeks before the index is published, it would be desirable for the unit to have more analytical capability. Moreover, the dependence of the CPI on a single trained statistician makes its compilation very vulnerable. In recent years, however, the Prices Unit has been assured of a core contingent of qualified staff. Their wages are considered acceptable when the bonuses allocated to enumerators are taken into account.

Enumerators in the Prices Unit are mostly suitably qualified (BAC + 2); the unit head is a graduate statistics engineer. There is no official program for theoretical and practical training in the CPI methodology and compilation methods. Because of staffing shortages, it is difficult for staff to take time off for training purposes. The head sometimes has the opportunity to attend meetings held by AFRISTAT or the WAEMU. Meanwhile, staff are not informed about the current revisions of the price manuals by the International Labour Organization (ILO) and other international organizations (in fact, the Prices Unit does not have a copy of the ILO manual on the CPI).

Computer resources are insufficient for adequate storing of data, and the capacity of PCs is such that freeze-ups may occur when the CHAPO software for compilation of the CPI is run. The current version of CHAPO, used in all the WAEMU member countries, needs to be improved (with respect to the correction of errors, adaptation to a new version of Windows, flexibility in application, etc.).

WAEMU financing of CPI current expenditure has ceased since December 2000. As of that date, these costs have been covered by the INSD budget. The release of funds sometimes occurs only a few days after the beginning of the reference month, which may affect the accuracy of the index.

0.2.2 Measures to ensure efficient use of resources are implemented

The INSD does not have specific methods and procedures in place to ensure the efficient use of resources.

0.3.1 Processes are in place to focus on quality

The INSD has not put a formal process in place to ensure a focus on quality (e.g., the concept of total quality or ISO 9000). External audits of the processes used can be carried out within the framework of technical or financial assistance projects. The INSD has not adopted a mission statement that stresses the importance of quality in all its dimensions.

0.3.2 Processes are in place to monitor the quality of the collection, processing, and dissemination of statistics

At the national level, there is no body distinct from the INSD that provides advice on the quality of the statistical series and on strategies for improving data production. Such a role could be assumed by the CNCS Methodology Committee, however, once it becomes active. To some extent, external monitoring takes place in the form of committees that comprise all the agencies involved in the production or use of certain statistics or groups of statistics.

Monitoring can also be organized at the regional level, for example by the Monitoring Committee for the Harmonized CPI (HCPI), members of which are the WAEMU, the BCEAO, AFRISTAT, and representatives of the national statistical agencies.

The INSD appears limited in its choices among the various dimensions of quality. It seems that the accuracy of the statistics is often given precedence over their accessibility and serviceability.

The INSD does not carry out user surveys and has not used any other means to obtain the opinions of users on data quality issues.

0.3.3 Processes are in place to deal with quality considerations, including tradeoffs within quality, and to guide planning for existing and emerging needs

The CNCS, which is intended to bring together producers and users of data, should review the existing statistical series and identify any new data needs in the context of the preparation of the annual statistics program. However, the CNCS is not in operation, and the INSD does little to encourage users to give their opinion on the statistical series. An exception is the foreword of several INSD publications, in which the reader is invited to make suggestions for improvement of the publication.

1.1.1 Statistics are compiled on an impartial basis

The Decree of 2000 setting up the INSD as an administrative government body states that the INSD is to be administered by a Board of Directors. The Chair of the Board is appointed by the Council of Ministers, and the other members of the Board are representatives of ministries, INSD workers, the Chamber of Commerce, and the BCEAO.

Although the Decree stipulates that the INSD Director-General has very broad powers to act on behalf of the Board of Directors, and in particular appoints and revokes personnel and proposes wages, the Decree contains no provision to the effect that the INSD is to be professionally independent or that external interference with respect to the content or the publication of statistical data is prohibited. Nevertheless, INSD management feels that the INSD can operate with full professional independence.

Professional skills play a decisive role in recruitment practices. Internal promotion is based on a rating system used in the civil service. Possibilities for promoting professionalism through the publication of methodology documents or participation in conferences, symposiums, and meetings is limited by financial constraints. Analyses carried out by the INSD for publication purposes are subject to internal procedures aimed at safeguarding the institution’s reputation for professionalism.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

Statistical considerations guide the INSD’s choices of source data and statistical techniques, although these are severely limited by the dispersed structure of the national system for the production of statistics and by financial constraints. Some choices are made at the regional level. For example, the methodological specifications for compilation of the CPI were established within the harmonized WAEMU framework. Both this methodology and the technical assistance received from AFRISTAT are informed solely by statistical considerations.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

Although management stated that it will take an active position if such is called for, there is no recent record of the INSD publishing comments on erroneous interpretation or misuse of its statistics. The INSD rarely provides explanatory documents or other information to the press and the other media with a view to preventing erroneous interpretation.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The laws governing the conditions under which statistical data are collected, compiled, and disseminated have all been published in the official gazette (JO) and are accessible to the public, including on the Internet at http://www.legburkina.bf/jo.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

There is no access for government employees to the HCPI data before their release. However, upon request the INSD can provide provisional estimates that are still subject to change. The BCEAO is the only institution that routinely receives the provisional figures, and this is indicated on the Burkinabè page on the IMF’s Dissemination Standards Bulletin Board (DSBB).

1.2.3 Products of statistical agencies/units are clearly identified as such

All INSD publications are clearly identified as such and also mention the ministry to which it belongs. Whenever INSD statistics are used or reproduced by other institutions (in practice, especially the BCEAO, WAEMU, and AFRISTAT) the source is always specified, as required by the INSD.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

Notice of major methodology changes is generally given to the public only at the time the new data are first disseminated. When the harmonized CPI was introduced, the INSD published a three-page leaflet describing the new methodology.

1.3.1 Guidelines for staff behavior are in place and are well known to staff

Behavior guidelines are the same for all government employees and are set forth in Law 010/98/98/AN of April 21, 1998 on the overall government reform.

In theory, INSD staff members take oaths in conformity with the provisions of the Law on statistical secrecy. However, these guidelines are not meticulously applied.

Ethical guidelines are not routinely explained to new employees on their arrival; nor are they periodically restated to personnel.

2.1.1 The overall structure in terms of concepts and definitions follow internationally accepted standards, guidelines, or good practices

The CPI methodology, concepts, and definitions are those agreed upon for the HCPI in the WAEMU member countries. In their overall structure, they are consistent with ILO guidelines. There are plans to develop indices according to secondary classifications (that is, by producing industries, origin, etc.) in the short run.

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices

CPI coverage is limited to Ouagadougou city, which is home to about 8 percent of the total population of Burkina Faso. In the medium term, the INSD hopes to be able to publish regional indices, as well as a national index.

The reference population of the current index is made up of African households, excluding non-African households, expatriate African households a member of which is employed by an international organization or the diplomatic corps, and collective units (such as boarding schools, hospitals, and prisons). Subject to these limitations, all households are covered, whatever the number of persons they comprise, the level of income, or main activity.

The transactions of the household sector covered by the CPI are a selection of “final consumption expenditure” as defined by the System of National Accounts 1993 (1993 SNA).

• By international convention, the index does not cover gross formation of fixed capital transactions (purchases of housing or machines, etc.); income distribution transactions, such as interest paid, income taxes, social contributions, pensions, or cash transfers to other households; and financial operations.

• Unlike in the recommendations of the 1993 SNA, but still in accordance with generally accepted practices, HCPI transactions are limited to monetary purchases. They do not cover all the goods (especially agricultural goods) produced by households for their own consumption. Also excluded are housing services that households produce for their consumption by virtue of the fact that they reside in their own homes. Accordingly, only the real rents paid by tenants are taken into account (about one-quarter of all housing units). This is acceptable according to international standards if the purchases of owner-occupied dwellings are included, which is however not the case.

• The coverage excludes, largely for pragmatic reasons, a number of goods and services such as health insurance, accident insurance, ancillary education services, and certain goods for leisure and culture (flowers, pets, etc.). Used goods that households sell to one another, such as second-hand automobiles and used clothing, are not covered in the index. However, when second-hand or used goods come from other sources, for example second-hand clothes from abroad, they are included.

The methodological documents available to the public clearly define the coverage differences between the CPI and end-consumption expenditure and the formation of fixed capital of the households sector, as defined in the 1993 SNA.

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices

The classification followed by the CPI is common to the WAEMU countries. The West African consumption classification (Nomenclature de Consommation Ouest Africaine—NCOA) is derived directly from the Classification of Individual Consumption by Purpose (COICOP) in the 1993 SNA. The NCOA has a structure that comprises 10 functions, 32 groups, 73 subgroups, and 105 items.

2.4.1 Market prices are used to value flows and stocks

The valuation rules used for recording expenditure and prices are consistent with the recommendations of the 1993 SNA. Final consumption expenditure is valued at purchasers’ prices, which includes all applicable margins and fees.

The specifications of varieties selected for the CPI include the characteristics that have an impact on their price.

2.4.2 Recording is done on an accrual basis

The prices of goods and services are recorded in the period that they are acquired. This rule is also followed if the amount of the expenditure is paid in several installments (e.g., in the case of a credit purchase).

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices

Problems concerning grossing/netting procedures are mainly related to the treatment of second-hand goods. The 1993 SNA defines consumption expenditure of households as purchases less sales. As the CPI does not cover used goods that households sell to one another (see 2.2.1), the grossing/netting problem does not arise. However, the scope of the CPI does include purchases of used goods from abroad without including sales of used goods to other countries; this introduces an inconsistency in the concepts of the CPI and a violation of the correct grossing/ netting recording procedure.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

Estimates of expenditure ratios for the CPI are based on the Ouagadougou household expenditure survey that was conducted during March 12–June 6, 1996, following a common WAEMU methodology. It is preferable that CPI weights are based on a survey covering a full year to avoid biases attributable to seasonal changes. On the other hand, the coverage of households by the 1996 survey was good. The primary units in the survey were areas assumed to contain 50–150 households each, as identified in a “Priority survey” ² carried out in 1994. A random selection was made of 84 areas from a total of 1204 areas. A count was taken in February 1996 of the households in the selected areas. To take account of extreme cases, nine households were selected in the three smallest areas, 15 households in the three largest areas, and 12 households in the other areas. In all, 1008 households were selected.

The sample households were grouped into four waves of 252 units each. Each wave was surveyed over an 18-day period, during which the enumerators visited the households every three days. The questionnaires contained several sections, including one on current expenditure recorded with the help of a diary, and other sections containing retrospective questions on expenditure over the past 12 months. Expenditure data were collected to a level of detail below COICOP level 4. Cases of nonresponse were handled using adequate imputation methods. Other adjustments, for example for underreporting, were not made.

The method used made it possible to calculate sampling errors. The results of the surveys carried out in the WAEMU countries show considerable variations in the structure of household expenditures among capital cities. For example, expenditures on foodstuffs, beverages, and tobacco in Ouagadougou (33.9 percent of the total) are relatively very low. One explanation could be expenditure on food consumption outside the home, which is recorded in the Hotel-Cafe-Restaurant function. No in-depth study has been carried out to better explain the differences between member states.

The Ouagadougou household expenditure survey has not yet been repeated, and such surveys are not compiled according to a predetermined scheme.

The CPI basket is described in terms of 10 consumer functions, 32 groups, 73 subgroups, 105 items, and 320 varieties. This level of detail is sufficient. Prices are monitored at 282 sample points of sale, including 30 housing units. Varieties and points of sale were selected judgmentally, with account taken of the presumed volume of business. A selection was also judgmentally made of 10 housing units each in the low-income, middle-income, and luxury categories. The INSD did not have access to a list of samples covering retail business. Total quotations average 3,250 a month.

On the markets, 30 quotations are collected each month for homogeneous varieties of nonmanufactured goods. In this regard, the INSD has adopted a different approach for each market: the largest market is visited four times a month, while the other markets are surveyed three times, twice, or once a month. A single quotation is collected for manufactured goods on the markets. At all the other points of sale, four price quotes a month are collected, while rental prices are recorded quarterly. Rates are collected on the basis of administrative sources or after consultation with the agencies concerned.

Homogeneous products sold in nonstandard units are purchased. Products sold in nonstandard quantities such as bundles or heaps are weighted at the Prices Unit. On their pricing forms, the enumerators specify the detailed characteristics of nonhomogeneous products, such as brand name, volume, origin, etc. Once the selection has been made, the product is monitored at every visit. Although prices are not necessarily collected on the same day of the month (or week, as the case may be), the enumerators take care to always collect the prices of food product in the morning to prevent undue fluctuations in the recorded data.

The need has not yet been felt to conduct ad hoc surveys to supplement the regularly conducted surveys. The press is not monitored as a source of information on items to be taken into consideration in the price collection.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

Source data are consistent with the definitions, scope, and classifications required for CPI needs. As regards the time of recording, all prices are collected during the reference month, except for rental prices, which are recorded on a quarterly basis, and rates. For some rates, notice is received at the times they change.

3.1.3 Source data are timely

The price survey timetable makes it possible every month to compile the CPI on the 10^th day of the month following the reference month.

3.2.1 Data compilation employs sound statistical techniques

The level of product detail selected is sufficient for analysis of price trends at the one-digit COICOP level. The monthly publication of the INSD also includes indices for unprocessed grains, meat, and fresh vegetables, as well as the average prices for 14 high-consumption products.

The CPI uses an unbiased formula to calculate elementary-level indices. This calculation is based on ratios of average prices with respect to homogeneous varieties, and as the average of indices for heterogeneous varieties. The approach adopted for aggregating the elementary indices is the standard version of the Laspeyres formula. This formula is generally accepted internationally, even if the revised Consumer Price Index Manual, which currently is being prepared by a group of international organizations under the chairmanship of the ILO, expresses a preference for other formulas.

The weight and index price reference periods date back to 1996. This is still acceptable, but they risk becoming outdated soon.

CPI weights are derived from a survey that covered only a few months of 1996, whereas the price reference period is the whole of 1996. The INSD has correctly adjusted the prices established on the basis of the survey, to align them with the price reference period. Due to lack of financial resources, the INSD has not been able to realize a survey covering 12 months, which would have guaranteed more realistic weights for the CPI.

After the 1996 weights were adopted, time series (only at the aggregate level) were compiled through the linking of the former index with the new one. In this procedure, 1997 was used as the overlap period for calculation of the link factors.

In the event that a product is temporarily unavailable, its price is assumed to be unchanged for a period of up to three months. After this three-month period, a replacement product is sought. The difference between the price of the formerly used product and that of its replacement is assumed to represent a difference in quality (in practice, a price in the base period is attributed to the new product, with the historical change in the price of the formerly used product applied to the current price of the replacement). Another way of estimating the base price of the new product is to consider that the price change is proportional to the changes in certain specified characteristics of the product (weight, volume, surface, strength, etc.). In practice, entirely new goods and services (e.g., cellular phones) are not included in the CPI basket.

The number of seasonal products distinguished is very low (less than 10). When these are unavailable on the market, prices are imputed to them on the basis of the prices of related products.

When a rate changes in the course of a month, the CPI reflects the correct average price for that month by the application of weights to reflect the number of days during which the former rate and the new rate were in effect.

3.2.2 Other statistical procedures (e.g., adjustments and transformations, and statistical analysis) employ sound statistical techniques

The portion of the economy not taken into account in the routine calculation of the CPI is not covered sufficiently by other price indices. The main reason is that the deflators estimated in the national accounts context have been preliminary since 1994. It should be borne in mind that major factors not covered by the CPI are household expenditure other than that effected in Ouagadougou, the share of agricultural production that is self-consumed, and the services of owner-occupied housing (about three-quarters of all premises). However, in April 1998 the INSD started the collection of monthly prices in 10 main regional towns. The prices of about 150 product varieties are monitored. Although the list of products differs from one region to the next, efforts made to draw up a list of common products have been successful. In the absence of data that could serve to establish weights, no aggregate indices are calculated. Regional prices are published irregularly in Dispositif de Suivi des Prix dans les Régions Economiques (System of Price Monitoring in the Economic Regions).

3.3.1 Source data—including censuses, sample surveys, and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide planning

Information is available on sampling errors for the Ouagadougou household expenditure survey of 1996. Information on nonsampling errors, such as insufficient or excessive coverage or classification errors, and nonresponse rates are not published for this survey.

Regarding the price collection, sampling errors cannot be calculated, given that varieties and points of sale were selected judgmentally on the basis of turnover.

Price data are first checked manually by the official in charge of the collection. They are then cross-checked in the audit module of the CHAPO software, which detects aberrant data. The module checks for consistency over time and with price changes for related products. Price Unit staff review the prices that CHAPO indicates as being improbable, this concerns about 50 percent for food products and 10 percent for manufactured goods.

The proper conduct of the survey and of CPI compilation, especially during the initial period, was monitored by INSEE, EuroCost, and AFRISTAT experts, who also provided technical assistance for the exercise.

3.4.1 Main intermediate data are validated against other information where applicable

Given that there is no producer price index or other monthly price indices (except for the prices of individual products collected in the regions), intermediate CPI data cannot be assessed against other sources. Neither are the prices noted for the CPI compared with those collected by the BCEAO, as the latter are not made available to the INSD.

3.4.2 Statistical discrepancies in the intermediate data are assessed and investigated

The INSD analyzes any unusual changes in the index caused by significant fluctuations in specific sectors or particular sources. For example, with Afristat collaboration, the INSD has investigated and responded to external questions regarding a strange development in grain prices. Also, two trainees from the National School of Applied Economy in Dakar recently carried out an in-depth study of the indices for homogeneous products and secondary classifications.

WAEMU and Afristat staff have the authority to cross-check the source data and indices published. With the departure of the price specialists, however, WAEMU and Afristat monitoring has declined.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

Potential statistical problems (e.g., aberrant price survey data) are investigated but are not generally pointed out to users.

3.5.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

In conformity with international practices, CPI data are not revised, except perhaps to correct errors (such as typing mistakes). There has not yet been any revision of the harmonized CPI.

4.1.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored

The CPI is used by the government and the monetary authorities for economic and social policy purposes. It is also used by civil society and by donors. There are no routine arrangements for consultations with these users through meetings or polls. In principle, the CNCS could assume this role. Nonetheless, the practical utility of the CPI in meeting users’ needs have been an important consideration during the development and adoption of the harmonized CPI.

The head of the Price Unit sometimes has the opportunity to attend seminars on the harmonized CPI held by AFRISTAT or the WAEMU, where he can learn of the needs expressed by users of the index.

4.2.1 Timeliness follows dissemination standards

CPI data for the reference month are final on the 10^th day of the following month. The memorandum disseminating the data leaves the INSD before the end of the month, as recommended in the GDDS (this timeliness also is in conformity with the Special Data Dissemination Standard—SDDS). The INSD is currently setting up a web site, which will enable it to reduce the time taken to disseminate the CPI to the general public down to about 10 days after the reference month.

4.2.2 Periodicity follows dissemination standards

The periodicity of the CPI is monthly, in conformity with both the GDDS and the SDDS.

4.3.1 Statistics are consistent within the dataset (e.g., observation of accounting identities)

The estimates issued are consistent, in that the results obtained by aggregation of the categories of goods or services are the same, whatever the typology of aggregation used.

4.3.2 Statistics are consistent or reconcilable over a reasonable period of time

A consistent time series without a break has been available since January 1996, which should be considered adequate. The series of the former CPI have been retropolated, at the aggregate level only, till 1982.

The publication of the CPI is accompanied by a commentary prepared by the INSD. Significant fluctuations in the monthly time series, when they occur, are briefly explained.

4.3.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

There are no price statistics from other sources with which the CPI can be reconciled. The national accounts use CPI components for the deflation of household expenditure.

4.4.1 Revisions follow a regular, well-established, and transparent schedule

Although the authorities consider it desirable to revise the CPI base with a periodicity of five years, it has not been possible to follow such a schedule. A new Priority Survey for the period April–June 2003 has been launched; it is essential for this survey to be followed by a supplementary household expenditure survey for July 2003–March 2004. The budget for this additional survey has been allocated but not yet released.

4.4.2 Preliminary data are clearly identified.

Not applicable (CPI data are not revised).

4.4.3 Studies and analyses of revisions are made public

The introduction of the harmonized CPI did not lead to a publication measuring and explaining the amendments made on the formerly compiled index. In light of the major changes introduced through the adoption of the harmonized index, such a publication would be of little value. The new base is amply discussed in the INSD publication Analyse des Résultats de l’Enquête sur les Dépenses des Ménages [Analysis of the Household Expenditure Survey Data].

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

The INSD disseminates the CPI every month in the publication Indice Harmonisé des Prix à la Consommation pour les Pays Membres de l’UEMOA [Harmonized Consumer Price Index for WAEMU Member Countries], which gives the general level of the index, the ten functions, three subaggregates of the function “Food, beverages, and tobacco,” several indices following the secondary classifications (origin, durability, etc.), and the average prices for 14 high-consumption products in CFA francs. The tables are accompanied by comments on changes in the index and a number of charts illustrating the changes to facilitate user analysis.

5.1.2 Dissemination media and formats are adequate

The data are disseminated in the form of a four-page memorandum entitled Indice Harmonisé des Prix à la Consommation pour les Pays Membres de l’UEMOA. The data are supposed to be also published in other INSD periodicals, such as Annuaire Statistique du Burkina Faso [Statistical Yearbook of Burkina Faso], but in practice there are important delays in their publication. For this reason, dissemination to the public depends largely on publications by other agencies, such as the Tableaux de Bord de l’Économie Burkinabè [Burkinabè Economic Indicators] by the Committee on Forecasting, Economic Conditions, and Multilateral Surveillance or the web sites of the BCEAO, the WAEMU, and AFRISTAT.

Users can obtain access to recently disseminated data and longer time series. However, this possibility is not publicized.

5.1.3 Statistics are released on a preannounced schedule

The methodological documentation and GDDS metadata indicate that the CPI for the reference month is disseminated on the 10^th day of the following month.

5.1.4 Statistics are made available to all users at the same time

Final data are made available to all users at the same time through the dissemination of the monthly memorandum Indice Harmonisé des Prix à la Consommation pour les Pays Membres de l’UEMOA.

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available on request

On request to the Prices Unit, the public has access to the various levels of aggregation, but this possibility is not publicized.

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

A methodological guide published by the WAEMU Commission, entitled Indice Harmonisé des Prix à la Consommation dans les Pays Membres de l’UEMOA—Théorie et Pratique [HCPI in the WAEMU Member Countries—Theory and Practice] is the reference instrument on methodology for all eight countries. More specific publications include the above-mentioned Analyse des Résultats de l’Enquête sur les Dépenses des Ménages; the memorandum prepared by Mathurin Dembo TOE, Méthodologie de l’Enquête sur les Dépenses des Ménages en 1996 [Methodology of the 1996 Household Expenditure Survey]; and the one-page explanation entitled IHPC: A Quoi Sert-Il? Comment Est-Il Elaboré? [HCPI: What Purpose Does It Serve? How Is It Compiled?] issued by the INSD when the harmonized CPI was being introduced. CPI metadata are also published on the IMF’s DSBB, abeit only in English.

However, the INSD’s Annuaire Statistique du Burkina Faso does not include any methodological notes on the CPI.

5.2.2 Levels of detail are adapted to the needs of the intended audience

The metadata are available at various levels of detail (see 5.2.1).

5.3.1 Contact person for each subject field is publicized

The name of the contact person for the CPI is specified in the monthly memorandum, with mail, telephone, fax, and e-mail contact information. General information can also be obtained from the INSD Documentation Unit.

There is no recent documentation to educate users on the disseminated indices (e.g., in the form of brochures), and users are not polled for monitoring of the assistance provided to them.

5.3.2 Catalogs of publications, documents, and other services, including information on any charges, are widely available

A catalog of INSD publications is available, clearly showing their prices. The Sales Unit can help with the placing of orders.

Table 2.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Price Statistics (Consumer Price Index)

(Compiling Agency: National Institute of Statistics and Demography—INSD)

Table 2.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Price Statistics (Consumer Price Index)

(Compiling Agency: National Institute of Statistics and Demography—INSD)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			No law making the INSD responsible for compiling the CPI. No coordination with the BCEAO, which also conducts price surveys.
	0.2 Resources				X		Too few professional staff in the Price Unit. Computer capacity inadequate. There is sometimes no money to make purchases in the markets.
	0.3 Quality awareness			X			No explicit arrangements are in place at the national level to monitor quality.
1. Integrity
	1.1 Professionalism			X			No official requirement for the INSD to operate independently.
	1.2 Transparency			X			Major methodology changes are not pre-announced.
	1.3 Ethical standards			X			Standards are not routinely explained to new staff, nor are they routinely reminded of ethical rules.
2. Methodological soundness
	2.1 Concepts and definitions		X					There are plans to calculate indices for secondary classifications (short term).
	2.2 Scope				X		Geographical coverage is limited to the capital. Rental services for owner-occupied housing are not covered by the CPI.	Compilation of regional indices and a national index (medium term).
	2.3 Classification/sectorization		X
	2.4 Basis for recording		X
3. Accuracy and reliability
	3.1 Source data				X		CPI weights are based on a survey carried out over a period of only three months. Housing for which rents are monitored was chosen without scientific weighting. Price collection is sometimes delayed.
	3.2 Statistical techniques		X					Rewriting of the CHAPO processing software (medium 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
	4.2 Timeliness and periodicity		SDD S
	4.3 Consistency		X
	4.4 Revision policy and practice			X			The change of the base year, planned for every five years, has been delayed.	A survey is under way for updating the index weights.
5. Accessibility
	5.1 Data accessibility			X			Hardcopy publication occurs a few weeks after the index has been finalized. The availability of supplementary data is not announced.
	5.2 Metadata accessibility			X			Secondary publications, such as the Annuaire Statistique du Burkina Faso, do not contain any explanations on the index.
	5.3 Assistance to users		X

View Table

Table 2.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Price Statistics (Consumer Price Index)

(Compiling Agency: National Institute of Statistics and Demography—INSD)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			No law making the INSD responsible for compiling the CPI. No coordination with the BCEAO, which also conducts price surveys.
	0.2 Resources				X		Too few professional staff in the Price Unit. Computer capacity inadequate. There is sometimes no money to make purchases in the markets.
	0.3 Quality awareness			X			No explicit arrangements are in place at the national level to monitor quality.
1. Integrity
	1.1 Professionalism			X			No official requirement for the INSD to operate independently.
	1.2 Transparency			X			Major methodology changes are not pre-announced.
	1.3 Ethical standards			X			Standards are not routinely explained to new staff, nor are they routinely reminded of ethical rules.
2. Methodological soundness
	2.1 Concepts and definitions		X					There are plans to calculate indices for secondary classifications (short term).
	2.2 Scope				X		Geographical coverage is limited to the capital. Rental services for owner-occupied housing are not covered by the CPI.	Compilation of regional indices and a national index (medium term).
	2.3 Classification/sectorization		X
	2.4 Basis for recording		X
3. Accuracy and reliability
	3.1 Source data				X		CPI weights are based on a survey carried out over a period of only three months. Housing for which rents are monitored was chosen without scientific weighting. Price collection is sometimes delayed.
	3.2 Statistical techniques		X					Rewriting of the CHAPO processing software (medium 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
	4.2 Timeliness and periodicity		SDD S
	4.3 Consistency		X
	4.4 Revision policy and practice			X			The change of the base year, planned for every five years, has been delayed.	A survey is under way for updating the index weights.
5. Accessibility
	5.1 Data accessibility			X			Hardcopy publication occurs a few weeks after the index has been finalized. The availability of supplementary data is not announced.
	5.2 Metadata accessibility			X			Secondary publications, such as the Annuaire Statistique du Burkina Faso, do not contain any explanations on the index.
	5.3 Assistance to users		X

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

According to the SP-PPF, satisfactory arrangements are in place to facilitate cooperation between the SP-PPF and the institutions and individuals responsible for financial statistics and the national accounts.

Thus, the SP-PPF works closely with the BCEAO. It sits on the Balance of Payments Committee, where balance of payments data are compared with those of the TOFE. With respect to foreign debt [drawings (in part), repayments, and interest], data used for the TOFE and the balance of payment statistics come from the same source: the DGTCP. Moreover, the SP-PPF is in constant contact with the BCEAO Research Department, which prepares the Net Treasury Position (PNT) and the Net Government Position (PNG) of the banking system. In practice, however, the SP-PPF does not seem to have sufficient and timely access to alternative and potentially better sources (monetary statistics) that could be used for statistical adjustments.

Cooperation with the producers of the national accounts (INSD) seems based on a practice from the time when GFS and national accounts were both produced by the Forecasting and Economic Conditions Committee in the context of producing the IAP (Automated Forecasting Tool) accounts (see in particular national accounts DQAF—element 3.2). In the wake of the organizational reforms of the Ministry, the SP-PPF participation to this Committee has yet to be formalized. Cooperation with INSD is limited for the time being to delivering TOFE figures for preparation of the (provisional) national accounts.

The National Statistical Coordination Council (CNCS), created in 1997 and installed in 2000, is currently inoperative (see national accounts DQAF—Dimension 0). Inter-agency coordination and liaison with users is provided in practice by the Forecasting and Economic Conditions Committee, which meets every four months to prepare the economic budgets, and by the National Economic Policy Committee (CNPE—Comité National de Politique Economique, for which the DGTCP is the Secretariat); these are the fora where GFS users (for the most part governmental) can express their needs.

The Budget Execution Monitoring Committee provides an opportunity to discuss new demands and to remind the different bodies of the need for coordination. The Macroeconomic and Budgetary Management Framework Group, in which the MFB, donors, the banking association and an association of nongovernmental organizations (NGOs) meet on a quarterly basis, is another opportunity for exchanging views on the TOFE.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

The principle of confidentiality applies only to the extent that, in compiling GFS, data may be obtained on individual nongovernment units. This is presently not the case in Burkina Faso where the GFS track the performance of government and not that of the public sector as a whole. SP-PPF agents, as public servants, are bound to observe confidentiality in the use of individual information in their possession.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

A 1995 circular, amended in 2002, establishes the duties of the various MFB departments to provide data to the SP-PPF, covering in particular:

• Tax collections, from the General Directorate of Taxation (DGI);

• Tax collections, from the General Directorate of Customs (DGD);

• Stages of the expenditure cycle, from the Computer Services Directorate (DSI) of the MFB;

• Correspondents’ accounts, Treasury receipts, movements of mandates/payment orders, domestic and foreign debt transactions and stocks, from the DGTCP;

• Operations (disbursements) financed by donors, from the General Directorate of Cooperation (DG-COOP);

• Reports of commitments by ministry and the “wage bill,” from the General Directorate of Budget (DGB);

• Report of commitments (aggregated), from the Central Financial Control Directorate (DCCF); and

• The PNT from the BCEAO.

The 1995 circular set a time limit of only five working days after the month in question for delivering the various statements. The TOFE working group meets around the 10th of each month to examine the draft monthly TOFE produced by the SP-PPF, and to pool information for identifying any corrections that might improve its overall consistency.

As regards public debt, the current information circuit is complex and does not allow for complete reconciliation of figures. While donors generally send due-date notices directly to the DDP, disbursements generally pass through the DG-COOP. Delays and a lack of detail have a significant impact on overall data reliability. Since 1998 the DDP has been conducting an annual survey of donors in order to reconcile data on the debt.

0.2.1 Staff, financial, and computing resources are commensurate with statistical programs

The staff complement (essentially two persons) available to the SP-PPF for preparing the TOFE is insufficient, in light of their other duties. This may reflect high turnover and the recent ministry reorganization. With the approaching departure of a resident, donor-financed consultant, there will be only one expert responsible for preparation of the TOFE, leaving the SP-PPF vulnerable. Moreover, the TOFE extension, called for in the WAEMU directives, and the SP-PPF increasing responsibilities for dissemination cannot be accomplished without more staff. Several positions in the SP-PPF organization chart are still vacant.

Computer equipment would seem to be adequate, but efforts are needed to make the TOFE promptly available via the MFB web site that is now in the process of being set up (or via alternative sites). SP-PPF officers have been trained at the Ecole Nationale des Régies Financières (ENAREF, the national financial management school for civil servants) and are given priority for training opportunities regionally or internationally: public finance courses (Washington); financial programming (Dakar); the GDDS seminar (2001, Bamako). For the most part, however, training is on-the-job. The SP-PPF expects to assign officers in charge of the TOFE the task of transmitting their know-how.

The other administration services involved in preparing the TOFE seem to have the necessary staff and consider themselves being reasonably well equipped.

The SP-PPF and the DGTCP are aware of the constant risk of losing their best officers to other organizations, and the disruption that this can cause. In this context, they believe that one form of incentive for motivating and retaining qualified personnel is to provide officers with high-quality training courses.

The debt management tool does not allow for managing stocks and flows within a single application, which would ensure consistency. Organizational shortcomings in terms of information circuit and qualified staff prevent the DMFAS (Debt Management and Financial Analysis System) software, in principle effective, from being used to its full advantage.

0.2.2 Measures to ensure efficient use of resources are implemented

The full costs of producing the GFS, in particular those of the TOFE, are not quantified or evaluated on a regular basis, because budgetary information on the SP-PPF and other source-data producers does not identify these costs separately. However, the detailed draft budget prepared by the SP-PPF quantifies some of the expenses (materials) relating to the TOFE.

Nevertheless, tasks and responsibilities would seem to be efficiently distributed within the MFB. With a view to tight production deadlines, the TOFE relies on existing data sources, which are only cautiously modified as new management tools are being introduced in the various offices: the main constraint on the use of source data has been the speed of delivery.

0.3.1 Processes are in place to focus on quality

Quality is considered the cornerstone of the statistical work that goes into preparing the TOFE. Improving the quality of the TOFE is one of the key points of the SP-PPF Activity Program. Nevertheless, insufficient account is taken of users’ needs: there should be easier access for the general public (the SP-PPF Activity Program calls for posting of the TOFE on the web site) and an obligation to explain methodology. Moreover, the GFS compilers have been reluctant to use standard statistical techniques to substantially improve the quality of information (grossing up, notably).

0.3.2 Processes are in place to monitor the quality of the collection, processing, and dissemination of statistics

The TOFE working group monitors the quality of information sources on a continuing basis. Sources are systematically compared, in particular with definitive Treasury sources and monetary statistics. The need to produce a satisfactorily balanced TOFE creates, de facto, conditions for a dialogue between SP-PPF and the producers of source data. On the other hand, the MFB seems to have a very narrow view of its dissemination responsibilities.

Source data rely largely on information reported by the Comptables publics (public pay officers), whose accounts are checked by the Audit Office (Cour des Comptes). In addition, the applicable financial regime sets strict conditions under which the pay officers (and also the ordonnateurs, or appropriation officers, and the financial controllers) must operate, particularly in terms of establishing accounting records. The authorities’ goal is to have a budget review law (loi de règlement de la loi finance) voted within two years. This assumes that the accounts are ready within a year for submission to the Audit Office, and then for vote by parliament. After a lengthy interruption, during which the last voted budget review law was for the year 1993, there has been some catching-up, as part of the structural adjustment program and as recommended by the ROSC on fiscal transparency. The latest budget review law voted is for the year 2000. The DGTCP believes that these new developments should produce a definitive cleanup of the Treasury’s ledger (BGCT); while its annual flows are satisfactorily recorded, the BGCT still suffers from provisional outstanding entries.

In preparing the TOFE, the SP-PPF follows the methodology in the IMF’s Government Finance Statistics Manual 1986 (GFSM 1986), in close cooperation with the African Department (AFR) of the IMF under the financial programs that have been in place with the IMF since 1991. In planning its statistical improvement program, the SP-PPF pays the greatest attention to the various WAEMU regional directives. Those directives cover:

• The budget laws (05/97/CM/WAEMU);

• The public accounts (06/97/CM/WAEMU and 05/98/CM/WAEMU and 05/99/CM/WAEMU);

• Budgetary nomenclature (04/98/CM/WAEMU and 04/99/CM/WAEMU); and

• The TOFE (06/98/CM/WAEMU).

The GDDS improvement plans also refer extensively to implementing these directives, which is planned for 2004 in the case of Burkina Faso. These improvements are being made in tandem with the statistical harmonization needed for the convergence programs (multilateral surveillance) in the WAEMU zone.

There are no specific provisions for soliciting the opinion of users, who often have difficulties linking the TOFE and the budget. WAEMU directives call for the TOFE to be presented as an annex to the budget law. The SP-PPF considers that it is willing to hear from users. Responding to donors’ requests, the SP-PPF is now reporting separately expenditures relating to debt relief under the Highly Indebted Poor Countries Initiative (HIPC).

0.3.3 Processes are in place to deal with quality considerations, including trade-offs within quality, and to guide planning for existing and emerging needs

Given the need to produce the TOFE within a very short time limit, for it to be useful as a budgetary monitoring and decision-making tool, the authorities have consciously opted for a TOFE that does not cover the entire general government and that is based on several sources instead of one single source: Treasury Ledger (BGCT—Balance Générale des Comptes du Trésor). Regular users are well aware of this choice.

Certain choices that the SP-PPF has made reflect a balanced approach to quality:

• Emphasizing the commitment basis for recording, which as the SP-PPF sees it, is the relevant basis for Burkina Faso. In the context of the WAEMU directives, the SP-PPF intends to move to a payment order basis, while retaining information on a commitment basis; and

• Keeping the monthly TOFE with a narrow coverage, while expanding it for the quarterly TOFE (TOFE-WAEMU).

Improving the quality of the TOFE is one of the action points of the SP-PPF Activity Program. Implementation of the integrated government accounts (see 3.1.1) and validation of the BGCT should allow compiling the TOFE using a single accounting source. Demands stem largely from the options adopted by the WAEMU. The TOFE has been adapted to meet the users’ demands for detailed data on HIPC expenditures.

1.1.1 Statistics are compiled on an impartial basis

The SP-PPF considers that it is recognized as the coordination body for preparing the TOFE. The TOFE relies largely on information recorded by public pay officers (comptables publics) in the course of their duties (but also by appropriation officers or ordonnateurs, and controllers), with the strict standards this implies. The SP-PPF believes that, because it reports directly to the Minister, it is sufficiently independent of the major MFB departments. The SP-PPF is well aware of the need for reprocessing statistics in order to move from a budgetary presentation to an economic presentation, and it feels free to deviate from the source data, if necessary.

The SP-PPF notes that the recent administrative reorganization has led to a certain slackening in analytical work, but the Activities Program deals clearly with this point. The SP-PPF recognizes that its officers will have to provide training for the major departments. The possibility of offering a TOFE course in the ENAREF is under consideration. The authorities are making use of regional technical assistance arranged by the WAEMU (with European Union financing) for implementing the TOFE-WAEMU.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

The SP-PPF feels obligated to follow the methodologies developed regionally or internationally. The WAEMU directives constitute a body of methodological references, the observance of which will be increasingly mandatory. The SP-PPF considers it is independent of any interference in the choice of sources and methods for the TOFE: for example, privatization below the line or the treatment of “compulsory loans.” It notes that the methodological choices made for the TOFE are thoroughly observed and discussed with AFR in the context of the structural adjustment program, and this reinforces the SP-PPF’s authority. This independence applies also in the case of donors, which sometimes favor one method of statistical recording (in the TOFE) for a transaction concerning them: for example, they may object to reclassifying “budgetary support” transfers under “project aid,” or to classifying transactions depending on the institutional sector of the project beneficiary.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

SP-PPF officers and senior officials regularly provide expert advice on technical aspects of the TOFE. However, they are not allowed to comment publicly on erroneous interpretation of the GFS, without express higher approval and, in the final analysis, without the permission of the Minister of Finance and Budget. The SP-PPF points out that this is a question of ministerial communications policy.

Little is provided by way of explanations in order to avoid erroneous interpretations. The SP-PPF and DGTCP recognize that the TOFE is still limited in its dissemination to a restricted circle of “initiated” experts, and that it is not yet a “consumer product.”

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

Decree No. 2002-466 of October 29, 2002, on the organization of the MFB, is publicly available, as are all organic laws, budgetary laws and various legislation organizing the government accounts and the budgetary nomenclature. These documents are available on the web site of the General Secretariat of Government. In addition, WAEMU directives are publicly available. On the other hand, the GFS dissemination practices and the role of the INSD are not made explicit and are therefore difficult for the public to understand.

Some of these elements are mentioned in the context of the GDDS metadata on the IMF’s Dissemination Standards Bulletin Board (DSBB), but only in English and not accessible from the ministry’s web site. The GDDS Table C Data Integrity and Access by the Public relating to GFS is not available on the DSBB. Restrictions on public access to the GFS are neither spelled out nor explained.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

To the extent that the TOFE is a statistical tool used for analysis in the conduct of economic policy, government officials who participate in economic and financial policy meetings have access to TOFE data before they are disseminated or published. In particular, MFB offices and the National Agency of the BCEAO have privileged access to the data. This fact seems to be publicly well known.

There are no rules governing this privileged access in terms of the approval procedures for publishing the GFS. The TOFE simply mentions the date on which “the data became available.”

1.2.3 Products of statistical agencies/units are clearly identified as such

The monthly documentation containing the TOFE and its annexes (which provide information often derived from other administrative units) carries the notation “SP-PPF” at the top of each page, without citing it as the “source.” It is best practice to make an entry below each table distinguishing between “source” (indicating the origin of the data) and “prepared by” (indicating the unit that has prepared the table itself and performed certain simple transformations: summation, the growth rate, data expressed as a percentage of GDP).

Tables published in the Burkina Faso Economic Report (2001) showing TOFE data cite the IAP (Automated Forecasting Tool) as source instead of the actual producer.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

Certain significant methodological changes have been notified at the time the new data were released, but not in advance: for example, the methodological changes relating to tax exemptions (1999 and 2001). Nevertheless, the move to the harmonized TOFE-WAEMU is the subject of fairly detailed directives. The associated methodological changes are thus implicitly announced in advance. The extension of the TOFE has been the subject of a detailed study that is publicly available, a fact that could be more widely publicized.

1.3.1 Guidelines for staff behavior are in place and are well known to the staff

There are no specific written rules concerning GFS. The staff, who produces the TOFE in the SP-PPF, and the public debt data in the DDP, operates in accordance with general administrative rules relating to professional misconduct and public service ethics (public service statutes: Law 13 on Public Employment). Also, producers of the TOFE and public debt statistics rely more generally on accounting records, the preparation of which is governed by strict standards, in light of the special status of DGTCP officers within the public service (the decree on the responsibility of public pay officers, or comptables).

The SP-PPF comprises officers who hold diplomas from the ENAREF, which offers a course in ethics. Most of them are also professional economists. The DGTCP and the DGD are both in the process of preparing ethical codes or guidelines. No such project is currently underway for the MFB as a whole, or for the SP-PPF.

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

Overall, the TOFE follows the structure and the methodology of the GFSM 1986: revenue, expenditure, lending minus repayments, from which a balance is derived, and financing with detail by counterpart (creditor/debtor). The debt tables cover the entire public debt, foreign and domestic. In some instances, the TOFE follows more closely recommendations of the new Government Finance Statistics Manual 2001(GFSM 2001) (commitment basis, privatization).

The general structure of the TOFE, as well as key elements of methodology, has been standardized for the WAEMU zone. Towards the end of the 1990s, the work of the Committee of Experts (WAEMU Secretariat) culminated in a series of WAEMU directives covering “government finance” aspects, and in particular the TOFE (see 0.3.2). The current practice of producing a TOFE that is already partially harmonized will become official upon implementation, in 2004 in Burkina Faso, of Directive 06/98/CM/WAEMU, which establishes standardization rules for the TOFE.

2.2.1 The scope is broadly consistent 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

The classification of revenue, expenditure, and financial stocks and flows seems to be consistent overall with the methodology in Chapter IV of the GFSM 1986. The boundary between “current transfers” and “goods and services” or between current spending and capital spending seems appropriate, except when it comes to externally financed projects. In this case, donors do not seem to provide statistical details concerning the economic purpose (investment, salaries, or goods and services) or the functional or sectoral distribution. Given the size of these operations, this represents a significant gap in the system.⁵ Meanwhile, the SP-PPF monitors the adequate sectoral classification of the beneficiary, and excludes from the TOFE any projects not for the benefit of general government.

Financial transactions, privatizations and loans, are correctly recorded below the line.⁶ The classification rule for financing is the residency criterion, while domestic financing distinguishes between the central bank, commercial banks and others. The classification criterion for the counterpart sector for negotiable securities is unfortunately the primary market (a very common error), which makes it difficult to compare TOFE data on financing with the monetary statistics and the PNT.

IMF debt is correctly classified in the TOFE under domestic financing (treated as an indirect loan through the Central Bank). However, some debt tables (stocks) show those amounts under foreign debt.

With respect to correspondent accounts, where the Treasury acts as a bank managing its clients’ deposits, the SP-PPF does not believe that the significant amounts recorded represent disguised quasi-budgetary activities or arrears management. Nevertheless, at the request of AFR, these amounts are included on an “adjustment to cash basis” line, before financing (except where they relate to banks and are appropriately recorded under bank financing).

With regard to the HIPC Initiative, by agreement with donors, debt service recorded in the TOFE is maintained unchanged. The difference to debt service actually due (and paid), which is the grant element under the HIPC Initiative, is recorded, at AFR’s request, on an adjustment line under financing, rather than under revenue. From an accounting point of view, the payment order in question mentions that only a portion is to be paid, the remainder being credited to the special HIPC account (a CST).

2.4.1 Market prices are used to value flows and stocks

Flows are valued on the basis of the amounts actually paid in terms of revenue, expenditure and financing. Domestic borrowing transactions are calculated from the change in stocks, which gives in Burkina Faso a satisfactory result. Interest on treasury bills (in the form of discounts) is recorded. Customs duties are recorded on an “accrual basis” (administrative concept). In practice, there are minor (and stable) differences between these amounts and those actually collected.

Gross debt is valued at face value.

Stocks of debt in foreign currency are converted, for the annual statement, at the rate reported by the BCEAO (on the last day of the period). Transactions in foreign currency use the effective rate (a BCEAO rate notified to DDP), in particular for drawings abroad, payments on the foreign debt, and budgetary support disbursements, an acceptable practice.

2.4.2 Recording is done on a cash basis (GFSM 1986 guidelines)

The TOFE records internally financed expenditure transactions on a commitment basis (base engagement), which the authorities believe is preferable because budgetary procedures in Burkina Faso give meaning to this time of recording. An adjustment below the line makes it possible to revert to the cash basis, which distinguishes expenditure committed but not mandated (DENO, Dépenses Engagées Non Ordonnancées) from changes in payment orders (instances de paiements). Such an approach already approximates a recording on an accrual basis, as called for under the GFSM 2001.⁷ Nevertheless, the SP-PPF plans to align its practice with the WAEMU directive and move to a payment order basis (Activities Program), while continuing to keep records of information on a commitment basis.

The TOFE correctly records revenues on a cash basis (except for customs duties, see 2.4.1). Data from the DGI are based on the accounting month (from the 25th of one month to the 24th of the next), and the SP-PPF occasionally makes adjustments where necessary.

The recording of externally financed projects presents a skewed time pattern over the course of the year, peaking in the final months. This reflects the fact that donor declarations are often late and information received from the DG-COOP does not allow transactions to be allocated accurately to each appropriate month.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices

The TOFE records transactions on a gross basis, with the exception of financial transactions which are recorded on a net basis, as is normal. Sometimes the TOFE helpfully identifies loan drawings separately from repayments. Sales by administrative entities are recorded as revenues. Expenditures are not netted with duties and other assigned revenue or taxes.

Since 1999 and 2001, tax exemptions relating to externally financed projects and oil supplies for the electricity company, respectively, are booked on a gross basis, as expenditure and revenue. Taxes collected on behalf of third parties (for example for local governments) are not recorded, which is normal. Bonuses distributed to collection officers (customs, taxes and treasury), particularly on penalties for late payment of taxes, are not recorded in the TOFE.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

A comprehensive data collection program is in place for GFS source data collection, taking into account Burkina Faso’s specific situation. As the Treasury ledger is not yet regularly produced, the source data for preparing the TOFE relies on reporting to the SP-PPF, as provided by a 1995 circular (see 0.1.4). The data is produced in the context of the internal management of departments/directorates involved in budgetary execution.

The DGTCP runs the network of pay offices ⁸, which keep records of revenues and expenditures.⁹ The aggregation of pay offices’ accounts results in the BGCT. Secondary pay offices act as collection offices for customs duties or taxes, or as accountants for the EPAs or local governments. Until the BGCT is fully operational, each network (DGTCP, DGI, DGD) compiles and centralizes its own information, which serves as source data for the SP-PPF.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

Budgetary and accounting information is for the most part reflected in the TOFE, after the necessary reprocessing. Consistent with the French system, the budget nomenclature differs from the government chart of accounts. The budget nomenclature is largely consistent with the TOFE categories While the calculation of the TOFE from accounting or budgetary items is not yet automatic, it is prepared for the most part using source data available in the context of departments’ internal management. In the future, WAEMU directives call for making this calculation automatic.

The timing of recording and the valuation methods of source data are generally consistent with GFS concepts. Statisticians seem to be fully aware of the differences between the concepts governing source data and those applied in GFS. The TOFE has the same scope (sectorization) as the budget, which reduces reconciliation problems. Nevertheless, the differences between the TOFE and the budgetary aggregates are not sufficiently explained, and it seems that many users do not understand them. The WAEMU directives call for the TOFE to be annexed to the budget law, which would provide a useful clarification for users.

3.1.3 Source data are timely

Source data are timely. The 1995 circular requires source data to be delivered to the SP-PPF within five days, a tight deadline that supposes flexible and direct delivery from the source data producer to the SP-PPF, with some room for the delivery of provisional estimates. More recently, this timing has slipped to as much as a month, perhaps reflecting a top-heavy information circuit. The expenditure process stages are recorded in real-time.

A decree establishing the DGTCP requires that the annual BGCT be prepared within six months. The DGTCP has set itself a deadline of 45 days for producing the monthly BGCT.

3.2.1 Data compilation employs sound statistical techniques

The TOFE, with its narrow coverage, relies generally on amounts actually observed.

3.2.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

Apart from externally financed projects, few adjustments are necessary because the TOFE is based primarily on accounting sources. Some ad hoc reprocessing may be needed to restore a superior TOFE equilibrium when it is disrupted by a large but awkwardly timed operation (for example a large tax payment received on the 27th of the month). No attempt has been made, however, to correct the current underestimation of the provisional figure for revenues from certain taxes, but the amounts involved are reportedly low.

For the provisional TOFE, the initial coverage of only about 70 percent of project loans and 40 percent of project grants is not subject to grossing up. The time of recording during the year is therefore noticeably distorted. The provisional annual TOFE is also systematically underestimated by significant amounts (no grossing-up). The time of recording of externally financed projects is currently the moment of disbursement by the donor. As the stock of funds still available to project managers appears in the monetary statistics, actual expenditure should be measurable by GFS compilers. The use of statistical ratios to allocate externally financed expenditure by economic type would also significantly improve the TOFE.

There are discrepancies between the TOFE and the monetary statistics when it comes to commercial bank financing. The TOFE approach focuses (erroneously) on the holder at issuance for negotiable securities, whereas the monetary statistics source could readily be used. The SP-PPF points out, however, that this information is often only available late.

Finally, the quarterly foreign debt series are not reestimated backwards, after reconciliation with donors to establish end-of-year stocks. Because of this, any statistical discrepancy is reflected entirely in the final quarter.

3.3.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide planning

The reliance on various source data with limited scope reflects the tight TOFE production schedules. However, the balance of priorities appears appropriate. The early data used for establishing the TOFE are sufficiently accurate for preparing and analyzing budget policy.

The SP-PPF regularly reconciles the various source data supplied to it, with data extracted from the BGCT or the CID, or by comparing data on debt service from the DDP with those from the CID, or with BCEAO financing. The TOFE working group meets to examine and assess the TOFE and to explain the results. Other interdepartmental meetings, mainly held for examining the fiscal and budgetary situation and preparing short-term forecasts, are the occasion to review data in detail. These sessions could however be better documented.

The record on the stock of foreign public debt is compared each year against the stocks declared (through a survey) by donors. The provisional nature of the early data is indicated and revised data bear a reference to the “reconciliation ratio” (indicative of coverage).

3.4.1 Main intermediate data are validated against other information where applicable

The SP-PPF states that it is performing reconciliations, whenever comparable statistics are available elsewhere. However, few such comparable statistics exist. The definitive national accounts are five years late. The provisional IAP accounts reflect the TOFE.

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated

The SP-PPF judges the plausibility of revenue data by comparing them with economic developments, but notes also that differences in trends may reflect changes in the efficiency of tax administration or merely a statistical problem.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

Concerning central bank financing, the TOFE relies on information provided by the BCEAO. Meanwhile, it is difficult to reconcile TOFE data on financing from the commercial banks with the monetary statistics, because of project loans, of the sectorization of SONAPOST (Post Office of Burkina Faso, Poste du Burkina), and of the fact that Treasury securities are sectorized by purchaser at issuance. Nevertheless, some more detailed items are compared.

3.5.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

The SP-PPF declares paying attention to revisions. There are however significant shortcomings in its current archiving policy. The provisional monthly TOFE is not archived, and it is not possible to reconcile it with the definitive TOFE. This is a particular problem when it comes to foreign financing disbursements. Reconciliations have been performed between the provisional and definitive source data provided by DG-COOP, but this has not led to any change in the method for calculating the TOFE.

Each month, the DGI provides figures for every month from the beginning of the year. The revisions are minor, mostly resulting from processing of suspense accounts, but they are nevertheless reflected in the TOFE. The first estimate tends to be regularly underestimated, but this has not yet led to any statistical reprocessing.

4.1.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored

The TOFE is a statistical tool that is integrated into the forecasting process, and is closely linked to the budget laws and to the monitoring of budget performance. It seems well suited to the needs of the authorities. However, the needs of other users are only partially taken into account: the TOFE is made public, but it is not readily accessible and it contains no explanation. It is published much later in the Tableau de Bord [Economic Indicators].

The timeliness observed in producing the TOFE means that it plays a major role in the exchange of views between administrations, and in the dialogue with the IMF. The TOFE is a key statistical element of IMF programs. Moreover, in the context of multilateral surveillance governed by Regulation 11/99/CM/WAEMU on implementing the Convergence, Stability, Growth and Solidarity Pact, each member state’s TOFE is reported to the WAEMU Secretariat along with its quarterly National Economic Policy Committee report. Total public debt as a percent of GDP, an annual statistic that is harmonized in the WAEMU, is published by the WAEMU Commission Secretariat (Report on Execution of Multilateral Surveillance by WAEMU) as a primary convergence indicator.

The TOFE is integrated into the publicly available accounts produced in the context of the IAP. These accounts provide several national accounts aggregates from 1985 to 2006, i.e., they include forecasts as well as recent estimates. IAP includes not only the TOFE but also simplified accounts for other general government. The TOFE therefore seems well integrated with macroeconomic forecasting.

There is no real GFS dissemination to the general public, except upon request, although it is now agreed that both the TOFE and the Public Debt should be available on the web. More generally, the TOFE is still a product designed for the “initiated” reader. It suffers from a lack of transparency in terms of its link to the budgetary aggregates, a recurrent question raised by many users. In the context of the WAEMU directives, it is planned to attach a TOFE based forecast to the budget. This move will be reinforced by bridge tables between the new budget nomenclature and the new TOFE.

4.2.1 Timeliness follows dissemination standards

Generally speaking, the dissemination timeliness is well within the GDDS recommendations, given a broad interpretation of dissemination. It would even satisfy SDDS requirements.

The monthly TOFE is disseminated within the month (except when there are problems), which would satisfy the SDDS rule and significantly exceed the GDDS requirement of three months. Aggregated quarterly data on the TOFE are published in the Tableau de Bord with currently a delay of about one year.

The TOFE’s coverage is too narrow to satisfy the scope of the GDDS central government comprehensive framework, and all the more so the SDDS indicator on the annual accounts of general government as a whole. Nevertheless, IAP data provide some aggregates on overall general government, within very short delays.

Quarterly public debt data are disseminated within one month (provisional data), which falls well within the GDDS and SDDS standards. An annual public debt series will be published in the 2002 annual report of the DGTCP.

4.2.2 Periodicity follows dissemination standards

Periodicities are well within GDDS dissemination standards, and meet those of the SDDS.

The TOFE is disseminated on a monthly basis, which exceeds GDDS criteria (quarterly), and would meet SDDS requirements. The main components of the quarterly TOFE are published in the Tableau de Bord.

The TOFE’s coverage is too narrow to satisfy the scope of the GDDS central government comprehensive framework, and all the more so the SDDS indicator on the annual accounts of general government as a whole. Nevertheless, IAP data provide some aggregates on overall general government.

Total public debt is disseminated quarterly (monthly for the domestic debt), which exceeds the GDDS rules (annual), and would meet the requirements of the SDDS (quarterly).

4.3.1 Statistics are consistent within the dataset

The aggregates correspond to the details in the tables, although the mission noted some anomalies in the various debt tables reviewed.

The articulation between flows and stocks (i.e. the link between net borrowings and changes in stocks) is satisfactory overall. Calculation of domestic financing relies on the amounts of domestic contractual debt, although a questionable presentation has created misunderstanding with users. On the other hand, for the foreign debt, the heterogeneity of sources is too great to guarantee full consistency between the change in stocks and transactions (after taking account of revaluations due to exchange rate movements). Moreover, revision procedures allocate all discrepancies to the fourth quarter.

Revenue, expenditure and balances are in line overall with financing. For expenditures, the difference between recording on a commitment and a payments basis is clearly reflected in a below-the-line item “accounts payable/receivable.” Given the TOFE’s narrow coverage, no benchmarking is required for the moment, and no significant consolidation issues arise.

4.3.2 Statistics are consistent or reconcilable over a reasonable period of time

The discrepancies noticed by the SP-PPF with regard to the expected trend of the time series for revenue or expenditure would seem linked to changes in budget policy or the underlying economic situation.

The past time series are not reestimated backwards to avoid discontinuities due to methodological changes (privatization, tax exemptions). The few existing breaks in series are not identified. The SP-PPF feels that readers will nevertheless be sufficiently informed if they pay careful attention to TOFE annexes and to the legend of the series.

4.3.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

The TOFE is used to prepare the aggregates for the national accounts, although the data in question, published in the context of the IAP, are only “provisional” (covering 1998–2002). The IAP macroeconomic series reflect the TOFE and supplement it with data on “other general government.”

Financing data from the TOFE are consistent with the monetary statistics, when they relate to the position of the BCEAO. However, discrepancies appear with commercial bank financing to the extent that the TOFE approach focuses (erroneously) on the holder at issuance for negotiable securities, and does not take into account deposits held in the banking system by managers of externally financed projects. The TOFE relies de facto on a narrow measure of PNG (the PNTT: PNT-TOFE), which is not always understood by users. An alternative approach would be to increase the coverage of the TOFE (TOFE-UEMOA).

Concordance between the balance of payments and the TOFE is ensured though the fact that they rely on the same source data (DGTCP).

4.4.1 Revisions follow a regular, well-established, and transparent schedule

Since the TOFE is primarily “accounting” in nature, there is no real policy or schedule for revisions. The SP-PPF believes that users expect some necessary marginal revisions to be made, in addition to the release of statistics for the new period.

4.4.2 Preliminary data are clearly identified

Preliminary data are not clearly identified. With the exception of externally financed expenditure (which constitute a separate section of the TOFE that is relatively unrelated to other lines), revisions to the TOFE and the public debt are of modest scope. Consequently, preliminary and definitive statistics are sufficiently consistent to allow relying on these preliminary statistics for formulating and analyzing economic policies and for day-to-day fiscal management. The various data extracted from the annual TOFE published by the BCEAO, and in the other publications, do not rely on provisional statistics.

4.4.3 Studies and analyses of revisions are made public

There is no chronological series of revisions. No study of revisions has been made public.

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

The TOFE is presented in a way that identifies the principal aggregates and balances, and links them to the underlying detailed information. The annexes to the TOFE provide a broad range of supplementary information, but no analysis is offered. The information presented in the TOFE is sufficient to track preparation and execution of the budget, although only the “initiated” can really understand this link. This should change with the planned move to annexing the TOFE to the budget law, in response to the WAEMU directives. The TOFE is integrated into programming tools (see IAP).

The level of detail in the TOFE and its annexes, as well as that on the public debt, is adequate, and even generous on a monthly basis. The scope is somewhat narrow but satisfactory on a monthly basis. Moreover, some data from the IAP can be used to supplement the general government data for certain key items.

The SP-PPF says that the TOFE time series and most of the annexes are available upon request. However, it does not maintain files with long series for the annual or quarterly TOFE. Time series on the public debt seem not available to the general public.

5.1.2 Dissemination media and formats are adequate

The TOFE and the public debt figures are essentially disseminated upon request, which is inadequate. The TOFE dissemination list includes some 60 addressees (of whom about 20 are in government). The TOFE is also available on diskette (Excel). The SP-PPF notes that the public can ask to be added to the dissemination list.

There are no publications covering the GFS as a whole, as is recommended. Certain elements of the TOFE and of the public debt are published in:

• Rapport sur l’économie du Burkina Faso (2001) (“Economic Report on Burkina Faso”)—annual, Direction des Etudes et de la Planification (Directorate of Research and Planning of the Ministry of Economy and Development);

• Tableau de Bord de L’Economie Burkinabè (1st quarter 2002)—quarterly, INSD;

• L’annuaire statistique (1999) (Statistical Yearbook)—annual, INSD;

• Le rapport annuel de la DGTCP (Annual Report of the DGTCP);

• Report on Economic Budgets (IAP); and

• Notes d’Informations et Statistiques (NIS) (Information and Statistics Bulletin)—annual, BCEAO. The NIS appear very late (the most recent issue accessible to the public is May 2002, showing the TOFE for 2001).

In short, GFS are published only as part of a broader economic bulletin. Published data are late, incomplete and lacking in detail. The means of dissemination are thus inadequate.

Plans are underway to post the TOFE on the MFB web site; the SP-PPF is a member of the web site steering committee. This could be the occasion to compile a single document, the “GFS booklet” containing:

• summarized and detailed data on the TOFE;

• data on the debt;

• methodological elements;

• organizational elements (terms and conditions); and

• links to other useful sites: DSBB for the GDDS and the Secretariat of the WAEMU Commission.

5.1.3 Statistics are released on a preannounced schedule

The dissemination date for the GFS is not announced in advance. The GFS dissemination policies are only implicitly known.

5.1.4 Statistics are made available to all users at the same time

The SP-PPF states that dissemination of the GFS beyond the MFB is done on an egalitarian basis, but it recognizes that this fact is not widely known. The units in charge of dissemination generally have a mailing list (monthly TOFE, annual report of the DGTCP), and the general public has the same access to this information, upon request.

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request

The SP-PPF provides supplementary details to the public on request, but this is not clearly announced.

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

The metadata posted on the DSBB for the GDDS provide elements of information, but they are available only in English, and the public is at no time invited to consult them (for example through an Internet link from the ministry’s web site). The metadata should be made available in French. The metadata in English on the DSBB should be updated.

A methodological note on preparation of the TOFE in Burkina Faso, dated January 1993, has been issued by the Ministry of Finance and Planning (with German technical assistance from the GTZ). It contains several bridge tables. Information on the TOFE-WAEMU is available in a consultant’s report produced in September 2002 entitled: “Operational implications of integrating EPAs, local governments, pension schemes and national funds into the TOFE.”

Also, the new WAEMU directives are publicly available, and give a good overview of the various methodological elements of the TOFE-WAEMU (in particular the bridge table).

5.2.2 Levels of detail are adapted to the needs of the intended audience

The published technical descriptions of concepts, sources and methods do not relate very closely to current production practices, as would be needed for experienced users to assess properly the strengths and weaknesses of the GFS. Although those descriptions are sufficient to understand the framework of statistics production, the information is not gathered (apart from the GDDS) in any single document or in a format suitable for use by non-specialists.

5.3.1 Contact person for each subject field is publicized

A GFS contact person, whose name is posted on the DSBB for the GDDS, is available to provide prompt and competent assistance to the public. This person can answer purely technical questions but there is no provision in place to respond to misuse of the GFS or to sensitize users.

5.3.2 Catalogs of publications, documents, and other services, including information on any charges, are widely available

There is no catalog indicating the list of publications where the GFS can be found, apart from a list of INSD publications and an incomplete list of documents mentioned on the DSBB.

Table 4.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Government Finance Statistics

(Compiling Agency: Ministry of Finance and Budget)

Table 4.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Government Finance Statistics

(Compiling Agency: Ministry of Finance and Budget)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			Overall coordination responsibility for GFS is not assigned to a single unit (e.g. SP-PPF). Arrangements related to dissemination are unclear. Current arrangements to ensure appropriate statistical reporting for data on externally financed projects are not effective.	Assign, by way of administrative decision, GFS leadership to SP-PPF, notably regarding methodology, dissemination, archiving, keeping in mind sharing of responsibilities.
	0.2 Resources				X		The lack of adequate staff resources currently observed, due to high turnover and reorganization, may have potentially disruptive consequences.	Assign immediately an economist to GFS at the SP-PPF. Complement the Activity Plan with on-the-job training objectives.
	0.3 Quality awareness			X			Relative neglect for the needs of users: views of users are not systematically sought; insufficient awareness of their needs, in particular concerning the provision of explanations. Reluctance to apply necessary statistical grossing up techniques.	Complement the Activity Plan with an objective to be closer to users’ needs.
1. Integrity
	1.1 Professionalism		X
	1.2 Transparency			X			No document describing the terms and conditions of current arrangements is available to the public, notably on internal access (GDDS Table C is missing). Sources are often not mentioned in various GFS tables (and their annexes).	Indicate appropriate sources under each table. Assemble, in a single area (on the web site), explanatory documents on GFS production.
	1.3 Ethical standards		X
2. Methodological soundness
	2.1 Concepts and definitions		X
	2.2 Scope			X			Coverage of the TOFE is incomplete (for annual data). There is no COFOG breakdown of expenditure. Debt comprises guaranteed debt (in some tables).	Extend TOFE to general government (UEMOA directives) (2004). End 2003 produce breakdown of expenditure by ministries and functions.
	2.3 Classification/sectorization		X					Implement a breakdown of externally financed expenditure (2005).
	2.4 Basis for recording		X
3. Accuracy and reliability
	3.1 Source data		X					Implement UEMOA directives (BGCT based TOFE) and the Integrated Government Accounting project.
	3.2 Statistical techniques			X			Provisional data on externally financed projects are not grossed up. Data, available in monetary statistics, relating to externally financed projects and to commercial banks’ claims on government, are not exploited.	Implement grossing up on externally financed expenditure, and use monetary statistics more fully.
	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		Inexistent archiving policy prevents meaningful revision studies. The few existing studies are not used to inform statistical processes.	Archive the TOFE. Carry out yearly revisions studies.
4. Serviceability
	4.1 Relevance			X			Users’ needs not sufficiently taken into consideration: tenuous link between TOFE and the budget, lack of explanations.	Annex the TOFE to the budget law and disseminate a bridge table between the TOFE and the budget.
	4.2 Timeliness and periodicity		SD DS
	4.3 Consistency			X			Users have difficulties to reconcile the TOFE with monetary statistics and financing flows with change in debt.	Improve the TOFE annex relating to the PNG.
	4.4 Revision policy and practice			X			The revision practices lack transparency and users are poorly informed of such practices (on externally financed projects, in particular). Preliminary data are not clearly identified in many tables.	Indicate status of data in all tables. Explain the revision practice.
5. Accessibility
	5.1 Data accessibility				X		Dissemination is mainly on demand. Publication appears well after finalization of data. Times series are not easily available. Availability of additional information is not publicized. No analysis is provided when GFS are disseminated. There is no pre-announced release schedule.	Build-up a GFS fascicule (web) with time series of various periodicities (TOFE and debt). Add a few annexes to the TOFE and a short analysis.
	5.2 Metadata accessibility			X			Accessible documentation on methods is loosely related to current TOFE compilation and is spread over too many places. No explanation is accessible for a basic audience, other than GDDS (which is in English and with no internet link with the ministry web site).	Add methodological explanations to the GFS fascicule; review GDDS metadata.
	5.3 Assistance to users			X			The only contact persons mentioned are those on the DSBB. There is no catalogue except an incomplete list on the DSBB (no internet link).	Indicate name of contract person on the web (or on the GFS tables).

View Table

Table 4.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Government Finance Statistics

(Compiling Agency: Ministry of Finance and Budget)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			Overall coordination responsibility for GFS is not assigned to a single unit (e.g. SP-PPF). Arrangements related to dissemination are unclear. Current arrangements to ensure appropriate statistical reporting for data on externally financed projects are not effective.	Assign, by way of administrative decision, GFS leadership to SP-PPF, notably regarding methodology, dissemination, archiving, keeping in mind sharing of responsibilities.
	0.2 Resources				X		The lack of adequate staff resources currently observed, due to high turnover and reorganization, may have potentially disruptive consequences.	Assign immediately an economist to GFS at the SP-PPF. Complement the Activity Plan with on-the-job training objectives.
	0.3 Quality awareness			X			Relative neglect for the needs of users: views of users are not systematically sought; insufficient awareness of their needs, in particular concerning the provision of explanations. Reluctance to apply necessary statistical grossing up techniques.	Complement the Activity Plan with an objective to be closer to users’ needs.
1. Integrity
	1.1 Professionalism		X
	1.2 Transparency			X			No document describing the terms and conditions of current arrangements is available to the public, notably on internal access (GDDS Table C is missing). Sources are often not mentioned in various GFS tables (and their annexes).	Indicate appropriate sources under each table. Assemble, in a single area (on the web site), explanatory documents on GFS production.
	1.3 Ethical standards		X
2. Methodological soundness
	2.1 Concepts and definitions		X
	2.2 Scope			X			Coverage of the TOFE is incomplete (for annual data). There is no COFOG breakdown of expenditure. Debt comprises guaranteed debt (in some tables).	Extend TOFE to general government (UEMOA directives) (2004). End 2003 produce breakdown of expenditure by ministries and functions.
	2.3 Classification/sectorization		X					Implement a breakdown of externally financed expenditure (2005).
	2.4 Basis for recording		X
3. Accuracy and reliability
	3.1 Source data		X					Implement UEMOA directives (BGCT based TOFE) and the Integrated Government Accounting project.
	3.2 Statistical techniques			X			Provisional data on externally financed projects are not grossed up. Data, available in monetary statistics, relating to externally financed projects and to commercial banks’ claims on government, are not exploited.	Implement grossing up on externally financed expenditure, and use monetary statistics more fully.
	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		Inexistent archiving policy prevents meaningful revision studies. The few existing studies are not used to inform statistical processes.	Archive the TOFE. Carry out yearly revisions studies.
4. Serviceability
	4.1 Relevance			X			Users’ needs not sufficiently taken into consideration: tenuous link between TOFE and the budget, lack of explanations.	Annex the TOFE to the budget law and disseminate a bridge table between the TOFE and the budget.
	4.2 Timeliness and periodicity		SD DS
	4.3 Consistency			X			Users have difficulties to reconcile the TOFE with monetary statistics and financing flows with change in debt.	Improve the TOFE annex relating to the PNG.
	4.4 Revision policy and practice			X			The revision practices lack transparency and users are poorly informed of such practices (on externally financed projects, in particular). Preliminary data are not clearly identified in many tables.	Indicate status of data in all tables. Explain the revision practice.
5. Accessibility
	5.1 Data accessibility				X		Dissemination is mainly on demand. Publication appears well after finalization of data. Times series are not easily available. Availability of additional information is not publicized. No analysis is provided when GFS are disseminated. There is no pre-announced release schedule.	Build-up a GFS fascicule (web) with time series of various periodicities (TOFE and debt). Add a few annexes to the TOFE and a short analysis.
	5.2 Metadata accessibility			X			Accessible documentation on methods is loosely related to current TOFE compilation and is spread over too many places. No explanation is accessible for a basic audience, other than GDDS (which is in English and with no internet link with the ministry web site).	Add methodological explanations to the GFS fascicule; review GDDS metadata.
	5.3 Assistance to users			X			The only contact persons mentioned are those on the DSBB. There is no catalogue except an incomplete list on the DSBB (no internet link).	Indicate name of contract person on the web (or on the GFS tables).

0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

Burkina Faso is one of the eight member countries of the West African Economic and Monetary Union (WAEMU) that share a common currency (the CFA franc—CFA stands for “Communauté Financière Africaine”) and a central bank (the Central Bank of the West African States—BCEAO for “Banque centrale des États de l’Afrique de l’Ouest”).

The Research Department at BCEAO headquarters and the National Directorate for Burkina Faso (DNBF for “Direction nationale pour le Burkina Faso”) jointly collect and process the source data and disseminate the monetary statistics of Burkina Faso. However, this responsibility is not mentioned explicitly in the Statutes of the BCEAO. Only Article 70 of the Statutes (Title IV—Miscellaneous Provisions, Section 5—Monthly Survey and Annual Report) provides that, for each fiscal year, a report on changes in the “integrated monetary survey” (SMI for “situation monétaire intégrée”)¹⁰ of the West African Monetary Union (WAMU) must be presented.¹¹ Taken literally, Article 70 does not extend to monitoring separately the monetary situation in each WAMU member country. Similarly, Article 36 of the Statutes (Title II—Central Bank Operations, Section 6—Support Provided by the Central Bank to WAEMU Governments) mentions only “the general monetary and financial situation of the Union and its outlook for change.”

The monthly BCEAO statement for Burkina Faso is prepared cooperatively by the DNBF and BCEAO headquarters staff. Under the terms of Article 19 of the Treaty Establishing the WAMU and Article 69 of the Statutes of the BCEAO, the BCEAO “each month shall prepare a statement of its accounts, which shall be published in the official gazette of each of the countries participating in the management of the BCEAO. Furthermore, each month the BCEAO shall prepare a statement by agency on issuance of currency and its counterparts.”

The accounting statements of banking institutions (commercial banks and finance companies) located in Burkina Faso are collected by the DNBF. The WAMU chart of accounts for banks (PCB for “Plan comptable bancaire”) explicitly provides that the monthly accounting statements of banking institutions shall be sent to the main national office of the BCEAO, i.e., to the DNBF in Ouagadougou in the case of Burkina Faso. The SMI for Burkina Faso is calculated by the DNBF and validated by BCEAO Headquarters in Dakar.

The monetary statistics for Burkina Faso are disseminated by BCEAO Headquarters simultaneously with those of the other WAEMU member countries in a monthly publication entitled Information Note and Statistics (NIS for“Notes d’Information et Statistiques”), which is also posted on the BCEAO web site (www.bceao.int).

0.1.2 Data sharing and coordination among data producing agencies are adequate

The procedures for exchanging and coordinating data among the units of the BCEAO (among the offices of the DNBF and between the DNBF and BCEAO headquarters) are laid down in a series of memoranda which were not provided to the mission; however, the procedures in place appear rational and adequate.

The procedures for the submission of statements of banking institutions are codified in detail within the PCB in terms of content and structure (formats of statements and computer files).

The data supplied by the Postal Checking Center (CCP for Centre des chèques postaux) and the National Savings Bank (CNE for Caisse nationale d’épargne) do not pertain to guidelines edited by the PCB, but are submitted in the framework of an ad hoc bilateral agreement. The same is true for the data obtained from the Treasury (cash in vault and secured liabilities).

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

Within the DNBF, the Monetary Analysis Section (SAM for “Service des analyses monétaires”) of the Research Unit (SE for “Service des études”) is responsible for compiling monetary statistics. Headquarters sends the DNBF accounting statement to the SAM. The SAM accesses the statements of the deposit money banks in the database managed by the DNBF Credit Unit (SC), which is the sole recipient of all periodic documentation from banks and finance companies (“établissements financiers” in the WAEMU banking law).¹²

An aggregated accounting statement for all commercial banks ¹³ can be created automatically via the software that the SAM uses, based on individual returns collected by the SC (series of form DEC2000 and following). For purposes of control or analysis, the SAM can also obtain individual accounting statements from the same software.

The monetary statistics produced by the SAM follow the principles of statistical secrecy, i.e., only aggregate data on all deposit money banks are published, and transactions carried out by each reporting institution cannot be detected. Only the transactions of the central bank, the CCP, the Treasury, and the CNE are identifiable. Owing to the specific missions assigned to these institutions by the government, identification of their individual operations does not contradict the rules of statistical secrecy.

The accounting statements of banking institutions and the various statements that make up the monetary statistics are kept in electronic files that are accessible through the DNBF computer network. DNBF staff have a personalized user name and password to access the network. An accreditation system authorizes staff member access only to those areas of information that are needed to fulfill the staff member’s functions within the DNBF. A security perimeter surrounds the DNBF premises where the SAM and the SC are located. Individual access within the perimeter is strictly controlled, and requires presentation of a badge.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

Law 012/96/ADR of May 2, 1996 regulating the banks (banking law) applies to banks and finance companies.¹⁴ Article 41 provides that “banks and finance companies must, during the fiscal year, prepare statements at the intervals and on the terms prescribed by the Central Bank. These statements shall be submitted to the Central Bank and the Banking Commission of the WAMU.”

The statutory deadline for banking institutions to submit their monthly statements to the BCEAO is one month. In case of noncompliance, substantial and sharply progressive penalties are provided (see Article 54 of the banking law); ¹⁵ these penalties are applied in practice (institutions were penalized for delays resulting from Y2K issues).

Although the banking law does not apply to the Post Office (see Article 2), the CCP must provide the BCEAO with information deemed necessary for the performance of its duties (see Articles 42 and 43).

The information that the BCEAO may obtain from banks, finance companies, and the CCP covers more than the data needed to produce the monetary statistics (bank risk, unpaid checks and commercial paper, and any other payment-related events). Professional secrecy may not be invoked against the BCEAO or the Banking Commission.

0.2.1 Staff, financial, and computing resources are commensurate with statistical programs

The SAM is composed of one management-level staff member. This staffing level seems adequate in view of the work load. This management-level staff member has a master’s degree. One year of practical experience is required to attain an adequate level in the area of monetary statistics. In May 2003, the staff member assigned to the SAM had 10 years of experience in this function. Another SE staff member is also capable of producing monetary statistics in the absence of the individual assigned to the SAM; that individual has five years of experience in this area.

DNBF staff has access to training courses held by the Training Center for the Banking Profession and to IMF seminars and courses. The departments of the BCEAO responsible for preparing monetary statistics have begun a discussion regarding implementation of the Monetary and Financial Statistics Manual (MFSM) published by the IMF in 2000 (in 2001 for the French version). A workshop on the redefinition of monetary aggregates was held by the BCEAO’s Research Department in Dakar in April 2003; two DNBF staff members took part in the workshop, and the Statistics Department of the IMF sent an expert in monetary statistics.

The SAM is equipped with a PC connected to the DNBF computer network. Processing of source data and calculation of the monetary statistics are fully computerized using the Fisec-EPM application.¹⁶ This software, which was developed by the BCEAO, is identical for the national directorates of the eight member countries of the WAEMU and for the headquarters departments.

0.2.2 Measures to ensure efficient use of resources are implemented

The human, financial, and material resources allocated to the “monetary statistics” function of the DNBF are determined by BCEAO headquarters, which ensures the consistency of resource allocations among the national directorates of the BCEAO. The DNBF requested an increase in staff size, in view of the new tasks for which the national directorates will be responsible in terms of monetary and financial statistics.

0.3.1 Processes are in place to focus on quality

There should be no trade-offs between deadline compliance and the completeness of the data. In fact, the small number of reporting entities (central bank, CCP, CNE, Treasury, and seven deposit money banks) allows final monetary statistics to be calculated directly. Monetary statistics are not disseminated to the public in a provisional version; only a final version is disseminated.

The general organization of the process for preparing and disseminating monetary statistics is designed and controlled by BCEAO headquarters. During the process, data are circulated quickly among the units involved, always via electronic file transfer.

BCEAO staff, particularly management-level staff, are concerned about data quality, and numerous controls are undertaken as monetary statistics are being prepared.

0.3.2 Processes are in place to monitor the quality of the collection, processing, and dissemination of statistics

The process of preparing monetary statistics is stable and not especially complex; there are few concerns to be raised.

The work of each unit within the DNBF is subject to the close supervision of a senior staff member who is appointed by, and reports to, the National Director. Moreover, inspection missions are sent by the BCEAO headquarters to ensure that every DNBF unit applies the system-wide management rules and work procedures. In particular, such supervision entails checking timeliness and periodicity of data collection, compilation and dissemination of monetary statistics. In addition, external auditing of DNBF work is carried out periodically at the unit level by an independent auditor of national standing, generally a certified public accountant licensed in Burkina Faso and selected through international competitive bidding procedures.

There are no surveys or inquiries of users to obtain their opinions on data quality issues.

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

Within the BCEAO, working groups composed of staff from BCEAO headquarters and the national directorates are set up when changes are made in the methodology relating to monetary statistics. Presently, a working group of this kind is preparing a review of the PCB, which will have an impact on the content and format of the data from banks that are used specifically for monetary statistics. The DNBF is participating in this working group.

The BCEAO does not conduct surveys of external users in order to obtain their views (satisfaction index, suggestions for improvement). Outside users are not specifically invited to express their comments or wishes. However, the BCEAO is open to any request by outside users for information regarding monetary statistics.

1.1.1 Statistics are compiled on an impartial basis

The monetary statistics of Burkina Faso are prepared by the DNBF, which is dependent in all matters (missions, resources) on BCEAO headquarters. In this area, the DNBF is therefore independent from the national authorities. DNBF staff members are permanent employees and cannot be dismissed except for serious offense. They are recruited either by competition or review of documentation for all candidates by a committee. Various degree requirements are set depending on the type of position to be filled.

Staff members are encouraged to develop their professional skills. They may produce memoranda or studies, some of which are published in the “Studies and Research” section of the NIS. They are made aware of training opportunities offered to the DNBF, and have access to training, seminars, colloquia, and meetings with other experts.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

The source data from all reporting entities and obtained directly from the source are sufficient to produce all the outputs expected within the methodological framework. The techniques used to produce monetary statistics do not rely on particular statistical methods; in fact, the source data are obtained through exhaustive collection at each cutoff date of the accounting statements from all the institutions that are included in the current scope of Burkina Faso’s monetary statistics.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The SAM and senior management have never had to face such a situation. In 1997, the exclusion of liquidated banks from the scope of the monetary statistics was mentioned explicitly, thus preventing from erroneous interpretations of breaks in data series.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The Statutes of the BCEAO and the banking law are accessible on the BCEAO web site. The public does not have direct access to any methodological note. No publicity is provided regarding the general conditions that govern the collection of raw data, their confidentiality, the compilation of results, or the obligation to disseminate them.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

There is no internal governmental access to monetary statistics prior to their release. The compilation and approval processes of monetary statistics are under the entire and sole responsibility of the BCEAO.

1.2.3 Products of statistical agencies/units are clearly identified as such

The NIS that contain monetary statistics are published under the BCEAO name without explicit reference to the DNBF, or to the SE or the SAM. The GDDS metadata accessible in English on the IMF web site mention the units of the DNBF and BCEAO headquarters that are responsible for producing the monetary statistics of Burkina Faso.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

The DNBF states that the conversion to the methodology contained in the MFSM will constitute a major change, owing to the innovations to be introduced with respect to current practices, particularly with regard to expanding institutional coverage, the preparation of flow data, and recording of claims and liabilities on an accrual basis. The DNBF considers that this conversion should be reported in advance to users, through joint agreement with BCEAO Headquarters.

1.3.1 Guidelines for staff behavior are in place and are well known to the staff

BCEAO staff members are subject to professional secrecy, under penalty of the sanctions provided in the criminal code (Article 47 of the Statutes of the BCEAO). They cannot hold any interests in any company, particularly financial institutions (Article 48 of the Statutes of the BCEAO). BCEAO staff, particularly at the time of their recruitment, are made aware of their obligation to comply with these ethical rules, and they undertake to do so in writing.

Title II of the BCEAO Staff Regulations (version of February 2, 2002) sets out the duties, obligations, and conflicts of interest for staff members. Article 4 deals with the obligation of fairness, and specifies that duties must be performed with full independence from external authorities and organizations. Article 5 deals with the obligations of discretion and restraint: “Staff members are bound by professional secrecy and must exercise restraint and absolute discretion with regard to all matters relating to the activities of the BCEAO” … “Termination of their employment, for any reason whatsoever, does not release staff members from these obligations, and former staff members of the BCEAO must avoid any actions that may cause difficulties for the Bank.” Article 6 indicates that “staff members may not engage in any activity that is incompatible with the independence and impartiality required in their roles as BCEAO staff, unless a waiver is granted by the Governor.” Finally, Article 7 provides that “staff members may not obtain or receive any investment, interest, or compensation of any kind whatsoever, through work or advice, in a public or private enterprise… without a waiver granted by the Governor. Any professional activity on the part of the staff member’s spouse must be declared to the BCEAO, which shall take appropriate steps to safeguard its interests, as necessary.”

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

The monetary statistics are broadly consistent with the principles presented in the Guide to Money and Banking Statistics in International Financial Statistics (Guide). The BCEAO expects to take the first steps toward implementation of the methodology contained in the MFSM. More specifically, following a technical assistance mission by the Statistics Department of the IMF in May 2001 to the BCEAO headquarters, the BCEAO indicated that the first steps toward bringing the monetary statistics of the WAEMU countries into compliance with the MFSM would take place in 2002, and would be continued in 2003; however, a final deadline for completing the migration has not been set. In April 2003, a workshop attended by all the heads of monetary statistics departments in the BCEAO was held in Dakar; it was decided on that occasion that a revised schedule would be established for implementation of the MFSM methodology, confirming de facto the lack of specific results to date.

For now, the BCEAO has not prepared a document indicating the main points of departure between the current monetary statistics methodology and that in the MFSM.

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices

The depository corporations now included in the SMI are the central bank, the commercial banks, the CCP for demand deposits, and the CNE for passbook savings deposits. The accounting statements of the commercial banks cover the operations of their headquarters and branches located within the national territory.

The accounting statements that the BCEAO receives from finance companies (credit institutions not authorized to collect deposits from the public) are not used in the context of the monetary statistics, although they could be used to calculate a partial financial survey (because insurance companies would still not be included).

Similarly, the operations of the decentralized financial system (SFD for “Système financier décentralisé”), which is dominated by savings and loan cooperatives and associations (“other depository corporations” as defined in the MFSM), which have grown significantly in recent years, are not taken into account in the monetary statistics, even though these entities collect savings deposits similar to those received by commercial banks. However, the institutions that are part of the SFD are already required to produce accounting statements and annual statistics which they submit to their supervisory ministry.

In December 1993, the WAEMU adopted a rule that applies to the SFD. To ensure the implementation of the rule, on July 30, 1999 the BCEAO established the Mission for the Regulation and Development of Microfinance (MRDM for “Mission pour la réglementation et le développement de la microfinance”), which in February 2003 became the Directorate of the Decentralized Financial System (DSFD for “Direction du système financier décentralisé”). Among other roles, it supervises, coordinates, and monitors programs that support regulation of the SFD, and is involved in strengthening the capacity of decentralized financial institutions and their control structures. The DSFD is also responsible for streamlining and improving SFD management and control practices, and creating a database on decentralized financial institutions. The DSFD works with the SFD monitoring unit set up within the Ministry of Finance, which supervises decentralized financial institutions in each WAEMU country. In Burkina Faso, the SFD is governed by the following: framework law 59/94/ADP of December 15, 1993, implementing order 95–308/PRES/MEFP of August 1, 1995, and establishing decree in the form of order 097–115/MEF/SG/DGTCP/ DAMOF, of August 18, 1997.

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices

The assets and liabilities of depository corporations are broken down first on the basis of the residence of the counterparty. Then the assets and liabilities against residents are broken down using a sectorization that is not sufficiently fine. In fact, the SMI for Burkina Faso presents a net government position that covers all of general government, although only the central government should be included. Also, the transactions of commercial banks with other financial institutions are not broken down by subsector (insurance companies, other financial intermediaries, financial auxiliaries).

Passbook savings deposits at the CNE are included in narrow money, while passbook savings deposits at commercial banks are included in time deposits.

Repurchase agreements between depository corporations are treated as loans, not as sales of securities with a repurchase option; therefore they are handled in compliance with the MFSM.

SDR allocations are included under foreign liabilities, although they should figure in the capital account in accordance with the IMF recommendations applied in most countries.

2.4.1 Market prices are used to value flows and stocks

Assets and liabilities in foreign currencies are valued at the average market rate of the last day of the month.

Burkina Faso does not hold stocks of monetary gold directly. Assets of this kind are centralized at the BCEAO headquarters, and do not appear in the accounting statement of the DNBF and hence not in the monetary statistics of Burkina Faso.

Directives 2001–02 and 2001–03 establish the accounting rules for securities belonging to banks and finance companies. Investment securities are recognized at market value. Securities issued by connected enterprises, equity securities, and long-term investments (portfolio formed with a view to return) are recognized at the lower of the acquisition cost and use value. Investment securities that must be held until their maturity date are subject to straight-line depreciation over their lifetime for the difference between book value and redemption value.

Monetary statistics are prepared in terms of stocks only; flows between two consecutive reporting dates are not calculated.

2.4.2 Recording is done on an accrual basis

The PCB (Directive 94-07 concerning the recording of linked claims and debts, Article 2) indicates that “on each closing date, accrued interest with respect to claims and debts must be entered in the accounts and subaccounts for linked claims and debts provided for this purpose, from the moment that this interest is transferred to the profit-and-loss account.” In fact, accrued interest not yet due is not transferred to the profit-and-loss account; accordingly, it does not appear on the accounting statements used to calculate the monetary statistics. Interest due but not paid is linked with claims and corresponding debts.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices

Doubtful claims are reported net of provisions established to cover potential losses. This practice is not consistent with the recommendations of the MFSM.

The PCB (Directive 94–06 concerning the principle of the nonuse of offsetting) indicates that no offsetting may be performed between the assets and debts of different legal entities, between the assets and debts of the same legal entity expressed in different currencies or having different terms, between a term deposit and a partial and/or temporary advance granted against that deposit, between interest expense debited from and interest income credited to the account of the same legal entity, and between income and charges relating to repurchase agreements. These provisions are consistent with internationally accepted principles.

In the NIS, the SMI presents the net position of the government and net external assets. However, gross claims and gross liabilities are presented in detailed tables for these two aggregates (Tables 2.5.2 and 2.2.2).

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

The source data for the current institutional coverage of the SMI (see 2.2.2) are exhaustive. These data are submitted by all institutional units in the subsectors included in the SMI for Burkina Faso. The data correctly reflect the reality of the monetary and banking transactions of Burkina Faso based on accounting statements. The BCEAO has a specific chart of accounts. Similarly, the banks and finance companies located in the BCEAO member countries have the PCB supplemented by two documents: the Compendium of Directives Concerning the Recording and Valuation of Bank Transactions (Recueil des instructions relatives à la comptabilisation et à l’évaluation des opérations bancaires), and the General Measures Concerning Models and Rules for the Submission of Statements (Dispositions générales relatives aux modèles et aux règles de remise des documents de synthèse). The PCB has been enacted on January 1, 1996, but seems to be still not fully adopted by all entities subject to it. Some of them have chosen to keep their previous chart of accounts by adapting it to the PCB through the use of an integrated bridge table in between both systems.

The stocks of demand deposits of nonfinancial agents at the CCP and savings deposits at the CNE are obtained directly from these two institutions using ad hoc statements (DEC611 and CNE statements, respectively).

Directive 94–10 concerning the recording and valuation of securities held by banks and finance companies has been replaced by Directive 2001–02 concerning the recording and valuation of subordinated securities acquired or issued by banks and finance companies and by Directive 2001–03 concerning the recording and valuation of securities other than subordinated securities acquired by banks and finance companies.

At the seminar on the redefinition of monetary aggregates held by the BCEAO in Dakar in April 2003, one recommendation concerned the establishment of a steering committee that would draw up a list of all primary data required in connection with the implementation of the MFSM, with a view to taking those data into account in the review of the chart of accounts of the BCEAO and the PCB.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

Approximations are not used as a substitute for source data, except with respect to the amount included under the “Banknotes and Coins” liability item in the accounting statement of the national directorates. Owing to the free circulation of banknotes among the WAEMU countries, the banknotes paid into the windows of the BCEAO by commercial banks may have been issued, in proportions that vary over time and space, by any of the eight countries in the zone.¹⁷ Accordingly, for a given country, the deduction of the total stock of banknotes to be sorted by country held in the vaults of the BCEAO within the territory of that country from the total issued by that country will result in overestimating or underestimating currency in circulation within the country, depending on whether the country is an overall exporter or importer of banknotes. To reduce these errors in measurement, BCEAO headquarters staff perform a breakdown by country of the stocks of banknotes to be sorted by country that are held in each country in the BCEAO vaults, and attributes to each country in the area the banknotes that it has issued and are considered held in the BCEAO vaults of the other seven countries. This breakdown is performed by using keys based on the results noted in each country following the banknote sorting operations by country. As commendable as this practice may be, nonetheless it remains flawed. It would be appropriate to rectify national currency in circulation figures not only by applying the distribution keys to the stocks of banknotes to be sorted by country in the BCEAO vaults, but also by applying these keys to the banks’ vault cash, and to the large numbers of banknotes that are actually circulating in the economies.

The monthly accounting statements of the commercial banks and the attached tables are submitted electronically (diskettes) to the Credit Unit (SC for “Service du crédit”) of the DNBF. Consistency checks on the submission are run when the data are loaded into the computer system: e.g. general balance of the accounting statement, and consistency of the attached tables with the accounting statement. In addition, the SAM, which is responsible for preparing the monetary statistics, performs accounting controls (stocks of refinancing transactions with the central bank through reconciliation with money market transactions, stocks of syndicated seasonal credits through reconciliation with the statements kept by the lead bank) and plausibility checks (for each item in the accounting statement, any change greater than two percent against the previous month is checked through via telephone with the declaring bank).

3.1.3 Source data are timely

Most of the seven commercial banks in Burkina Faso submit their monthly accounting statements within the statutory 30-day deadline. One or two commercial banks are often late by one or two weeks. The SC, the recipient of these statements, issues reminders to late filers. These reminders may be in the form of letters of injunction sent to the managing directors of such institutions. The letters remind recipients of the financial penalties provided under Article 54 of the banking law.

The accounting statement of the DNBF is prepared by the departments of the BCEAO headquarters in Dakar. On May 12, 2003, the most recent DNBF statement available was dated January 31, 2003.

The data from the CCP (stock of demand deposits) are submitted in a timely fashion. The data from the Treasury (cash in vault and secured liabilities) and from the CNE (stock of savings deposits) are often submitted late, owing to the size of their network through the entire country, and the difficulty in maintaining ties with offices far from the capital; the most recent known data are reported if the DNBF has all the other information needed to calculate the SMI.

3.2.1 Data compilation employs sound statistical techniques

The procedures for collecting source data, including on diskette for banks and finance companies, allow for computerized processing. The Fisec-EPM database used jointly by the SC and the SE for preparation of the monetary statistics is managed by experienced staff. The successive aggregations and consolidations performed as the monetary statistics are compiled follow the principles of the major accounting identities.

3.2.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

The BCEAO does not publish seasonally adjusted statistical series. It does not make substitutions for source data, because the source data are exhaustive. Provisional carry-forwards of data are done to make up for delays on the part of smaller declaring entities (Treasury, CCP, CNE); when the final data are received, they are used to replace the data carried forward. However, after a two-month period, the data carried forward are no longer corrected.

3.3.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide planning

The source data used come almost exclusively from accounting sources. The reporting entities constitute the entirety of the population covered; sampling is not used.

3.4.1 Main intermediate data are validated against other information where applicable

The monetary statistics compilation process does not provide for producing intermediate outputs. Calculation of the main aggregates and their breakdown is based directly on accounting data; no other, more reliable sources with the same coverage exist.

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated

The compilation process for the monetary statistics does not provide for the production of intermediate data.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

In case of a significant or unexplained fluctuation in the monetary statistics, contact is made directly or by letter with the declaring institution that supplied the data that caused the fluctuation.

3.5.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

The monetary statistics compilation process does not include systematic revision. Monetary statistics are considered final, and only isolated corrections may result in a revision of the outputs. The compilation process is highly stable, and error corrections are infrequent.

4.1.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored

Each WAEMU member country has two representatives on the Board of Directors of the BCEAO (Article 49 of the Statutes of the BCEAO). Each member country also has a national credit committee supported by the National Directorate of the BCEAO and chaired by the Minister of Finance. These relationships between the BCEAO and the authorities enable the latter to express their views, as necessary, including on the relevance of the monetary statistics for the analysis of the country’s economic and financial position.

The DNBF maintains regular contact with institutional users and correspondents such as the SP-PPF (especially for the transition from the government’s net position to the net position of the Treasury, which has a more restricted coverage), the INSD, and the Directorate of Forecasting and Macroeconomic Analysis (in the context of the Committee for Forecasting and Multilateral Surveillance). However, the DNBF does not hold meetings or arrange contacts with other users. The DNBF takes part in the “African Statistics Day” held each year, but does not make a special presentation at it.

The BCEAO integrates into its methodology for producing monetary statistics the consequences of certain changes that affect the financial sector. A revision under way of the PCB, for example, aims to take into account the economic nature of financial leasing transactions, which will be followed by an adjustment of the monetary statistics.

4.2.1 Timeliness follows dissemination standards

On May 10, 2003, the statements of the central bank and deposit money banks posted on the BCEAO web site were those for the period ending in December 2002. With regard to the NIS, the most recent edition, published in April 2003, is dated “May 2002” and includes the two statements as at end-March 2002.

The GDDS recommends dissemination of the central bank sectoral balance sheet within two months following the end of the reference month. Accordingly, as of May 12, 2003, the January 2003 report is more than one month late on the web site, and is one year late in the NIS.

The GDDS recommends dissemination of the deposit money bank sectoral balance sheet within three months following the end of the reference month; accordingly, the deposit money bank sectoral balance sheet for January 2003 is ten days late on the web site and ten months late in the NIS. However, this sectoral balance sheet is available within the BCEAO, as are those for February and March 2003. These sectoral balance sheets of the deposit money banks could be disseminated by the BCEAO independent of the central bank sectoral balance sheet.

4.2.2 Periodicity follows dissemination standards

Monetary statistics (specifically the central bank and deposit money bank statements) are compiled monthly in conformity with the recommendation of the GDDS.

4.3.1 Statistics are consistent within the dataset

The internal consistency of the monetary statistics is assured; the various statements and detailed tables that make up the monetary statistics are carefully structured (See 5.1.1). Reciprocal assets and liabilities of the central bank and the other depository corporations are consistent. Flow data are not available, and thus the consistency between stocks and flows cannot be checked.

Because the other finance companies are not covered by the monetary statistics, the consistency of reciprocal assets and liabilities of depository corporations and other finance companies cannot be checked.

The BCEAO cannot reconcile the sectoral distribution of the components of the money supply and the data presented in the national accounts. In fact, Burkina Faso does not produce financial accounts or flow of funds. For the same reason, the credit aggregates of the monetary statistics cannot be reconciled with debt data from the financial account.

4.3.2 Statistics are consistent or reconcilable over a reasonable period of time

The monetary statistics are available as monthly data and in accordance with a consistent methodology used since 1994, through the database managed by the DNBF. In Excel files, BCEAO Headquarters has longer series covering the period from June 1985. A history covering the period from 1965 through 1984 is available from the Research Department at BCEAO Headquarters.

The main breaks in the time series are indicated (for example: starting in January 1997, the exclusion of previously liquidated commercial banks; however, the scope of the impact of this exclusion on the main series is not indicated).

4.3.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

The consistency between external assets and liabilities of the SMI and the international investment position is checked; a reconciliation table between these aggregates is included in the annual publication on the balance of payments statistics.

The deposits and liabilities of the central government with the central bank as they appear in government finance statistics are consistent with the comparable monetary statistics data; in fact, the government finance statistics use the monetary data. However, some divergences remain, resulting particularly from differences between the BCEAO and the Treasury regarding the institutional coverage of the state and central government concepts. Also, deposits and liabilities of the central government with other depository corporations as they appear in government finance statistics are not identical with the comparable monetary statistics data, but these discrepancies can be explained.

4.4.1 Revisions follow a regular, well-established, and transparent schedule

The structure of the monetary statistics compilation process does not include a revision procedure.

4.4.2 Preliminary data are clearly identified

The monetary statistics are calculated and disseminated in final form only; no preliminary version is produced.

4.4.3 Studies and analyses of revisions are made public (see also 3.5.1)

Procedures of this kind are not applicable with respect to the monetary statistics of Burkina Faso.

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

The monetary statistics of Burkina Faso are published each month in the NIS, under the BCEAO Headquarters’ name. The NIS brings together the national statistics of all the WAEMU member countries. In the NIS, one of the four sections presents the monetary and financial statistics.

The monetary statistics are comprised of seven tables of figures:

• survey of monetary institutions (integrated monetary survey—SMI),

• net position of government,

• central bank survey,

• banking survey (deposit money banks),

• net foreign assets of monetary institutions,

• uses of credit reported to the Central Risk Recording Service, and

• base money.

These tables of figures include data for the three most recent months and the data at end-December for the previous eight years.

Three charts present:

• the integrated monetary survey (net foreign assets, net position of government, money supply, domestic credit, credit to the economy),

• the money supply and three of its components (currency in circulation, money,¹⁸ bank deposits),

• credit to the economy (total credit, seasonal credits, short-term credits, medium- and long-term credits), and refinancing obtained at the central bank.

In addition to the national data of the eight member countries, the NIS presents the same data consolidated for the entire WAEMU area.

The NIS does not include commentary on changes in the SMI.

5.1.2 Dissemination media and formats are adequate

There is no national monetary statistics publication. No press releases are issued when the data are published. In addition to the publication of the NIS in printed form, the web site that BCEAO Headquarters administers also publishes recent monetary statistics. The online information is a facsimile of the tables of figures that appear in the NIS (charts are not included on the web site). The data posted on the web site are more up to date than those published in the NIS (see 4.2.1).

5.1.3 Statistics are released on a preannounced schedule

There is no preannounced schedule for the dissemination of monetary statistics.

5.1.4 Statistics are made available to all users at the same time

The National Economic Policy Committee, the Treasury, the STC/PDES, and the SP-PPF receive the monetary statistics prior to their dissemination to the public. This prior communication is not made known to the public; the GDDS metadata published on the IMF web site do not mention this practice. This communication generally occurs a few days after the monetary statistics are available. As a result, these recipients have the monetary statistics several weeks or months before the public, owing to the time lag in disseminating the data depending on the medium used (web site, NIS).

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request

Requests for nonpublished subaggregates are sent to the BCEAO headquarters, which has a database with longer series than those available to the DNBF. However, the DNBF still serves as the requester’s contact, and the response to the request is made through the DNBF.

The availability of nonpublished subaggregates is not publicized.

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

The main concepts and scope of the monetary statistics, as well as differences from international statistical standards, are described in the GDDS metadata, available in English on the IMF web site. The French version of the metadata is not presently available to the public in print or on the Internet.

The BCEAO has a 1995 methodological note for the monetary statistics, entitled “Compilation of Monetary Statistics by the BCEAO” (Central Research and Forecasting Department); the links with other economic statistics systems are not highlighted in the document. This methodological note, which dates back eight years, is not published or posted on the BCEAO web site.

The PCB can be obtained from the BCEAO for a nominal fee. By contrast, the BCEAO’s chart of accounts is not accessible to the general public, nor is it disseminated to interested users upon demand. Instead, this chart of accounts can be consulted only on the BCEAO’s premises by authorized individuals at the discretion of the National Director.

5.2.2 Levels of detail are adapted to the needs of the intended audience

All users receive the same set of data on monetary statistics. The only exception is the more detailed net government position-net Treasury position (with identification of each Treasury account) that is submitted to the authorities.

No methodological document has been developed targeted to an intended audience.

5.3.1 Contact person for each subject field is publicized

The staff of the SAM is available to internal users at the DNBF to provide any assistance needed with regard to the monetary statistics.

The NIS does not publicize any contact person for monetary statistics other than the director of the publication (Director of Statistics and Research—BCEAO Headquarters). Similarly, no regional or national contact for monetary statistics is identified on the BCEAO web site. Meanwhile, the GDDS metadata on the IMF web site identify contacts for each category and indicator of financial sector statistics (name, title, address, telephone, fax, and e-mail information).

5.3.2 Catalogs of publications, documents, and other services, including information on any charges, are widely available

The NIS includes a notice regarding how to obtain the publication (address, single issue price, subscription rate). The NIS does not mention the existence of the BCEAO web site.

The catalog of periodical and nonperiodical publications is available on the BCEAO web site. All the necessary information is provided concerning how to obtain and subscribe to the publication (contact, single issue price, subscription rate).

Table 4.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Monetary Statistics

(Compiling Agency: BCEAO - National Directorate for Burkina Faso)

Table 4.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Monetary Statistics

(Compiling Agency: BCEAO - National Directorate for Burkina Faso)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			The responsibility of the BCEAO for monetary statistics, particularly national monetary statistics, is not clearly established.
	0.2 Resources		X
	0.3 Quality awareness			X			A partial financial survey including finance companies is not established although primary data are available quarterly (previous TA missions in the zone made this recommendation in 2001). The inclusion of cooperatives and mutual saving and credit institutions (decentralized financial system) in the monetary survey is still pending although all directives for collection of primary data were set up five years ago.
1. Integrity
	1.1 Professionalism		X
	1.2 Transparency			X			No documentation on the terms and conditions under which statistics are collected, processed, and disseminated is available to the public.
	1.3 Ethical standards		X
2. Methodological soundness
	2.1 Concepts and definitions			X			The MFSM methodology requirements of processing data from financial institutions beyond depository corporations and organizing a comprehensive set of flow data to complement stock data are not met.	Following the workshop on the revision of the monetary aggregates definition held in Dakar during April 22–25, 2003, it has been recommended to set up a committee to design a coherent strategy aiming at the implementation of the MFSM by the BCEAO.
	2.2 Scope			X			Cooperatives and mutual saving and credit institutions (decentralized financial system) are not included in the monetary survey.	Inclusion of the finance companies in the monetary statistics; collection of data on transactions performed by the decentralized financial system for inclusion in the monetary survey.
	2.3 Classification/sectorization				X		General government institutions are not properly sub-sectorized in central administrations, local government, and social security institutions. Allocations of SDR are included in foreign liabilities instead of in capital accounts. Passbook savings with the Caisse nationale d’épargne (CNE) and passbook savings with the commercial banks are not included in the same sub-component of the money supply.	Improvement of the sectorization of general government units in the monthly reporting of commercial banks.
	2.4 Basis for recording		X
3. Accuracy and reliability
	3.1 Source data			X			Delays in establishment of monthly accounts of the BCEAO National Directorate do not allow for a dissemination meeting GDDS recommendations.	Implementation of the MFSM methodology.
	3.2 Statistical techniques		X
	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			No organized relations with external users, but methodological changes explained by compilers when deemed necessary.
	4.2 Timeliness and periodicity				X		Dissemination lags exceed GDDS recommendations, especially the printed monthly bulletin “Information Notes and Statistics” (NIS).	Renovation of the computerization of all the BCEAO’s accounting functions, reduction in the time taken to disseminate monetary statistics to 45 days.
	4.3 Consistency		X
	4.4 Revision policy and practice		X
5. Accessibility
	5.1 Data accessibility				X		Statistics are made available to selected users much before the public at large. No pre-announced release schedule. No comments are issued with the monetary statistics.	Publication of a preannounced release schedule.
	5.2 Metadata accessibility				X		No dissemination of methodological document to the public (apart from GDDS metadata in English on the IMF web site).	Dissemination to the public of the monetary statistics’ methodology.
	5.3 Assistance to users			X			In the NIS and on the BCEAO web site, no contact person is publicized; contact persons are mentioned in the GDDS metadata only.

View Table

Table 4.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Monetary Statistics

(Compiling Agency: BCEAO - National Directorate for Burkina Faso)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			The responsibility of the BCEAO for monetary statistics, particularly national monetary statistics, is not clearly established.
	0.2 Resources		X
	0.3 Quality awareness			X			A partial financial survey including finance companies is not established although primary data are available quarterly (previous TA missions in the zone made this recommendation in 2001). The inclusion of cooperatives and mutual saving and credit institutions (decentralized financial system) in the monetary survey is still pending although all directives for collection of primary data were set up five years ago.
1. Integrity
	1.1 Professionalism		X
	1.2 Transparency			X			No documentation on the terms and conditions under which statistics are collected, processed, and disseminated is available to the public.
	1.3 Ethical standards		X
2. Methodological soundness
	2.1 Concepts and definitions			X			The MFSM methodology requirements of processing data from financial institutions beyond depository corporations and organizing a comprehensive set of flow data to complement stock data are not met.	Following the workshop on the revision of the monetary aggregates definition held in Dakar during April 22–25, 2003, it has been recommended to set up a committee to design a coherent strategy aiming at the implementation of the MFSM by the BCEAO.
	2.2 Scope			X			Cooperatives and mutual saving and credit institutions (decentralized financial system) are not included in the monetary survey.	Inclusion of the finance companies in the monetary statistics; collection of data on transactions performed by the decentralized financial system for inclusion in the monetary survey.
	2.3 Classification/sectorization				X		General government institutions are not properly sub-sectorized in central administrations, local government, and social security institutions. Allocations of SDR are included in foreign liabilities instead of in capital accounts. Passbook savings with the Caisse nationale d’épargne (CNE) and passbook savings with the commercial banks are not included in the same sub-component of the money supply.	Improvement of the sectorization of general government units in the monthly reporting of commercial banks.
	2.4 Basis for recording		X
3. Accuracy and reliability
	3.1 Source data			X			Delays in establishment of monthly accounts of the BCEAO National Directorate do not allow for a dissemination meeting GDDS recommendations.	Implementation of the MFSM methodology.
	3.2 Statistical techniques		X
	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			No organized relations with external users, but methodological changes explained by compilers when deemed necessary.
	4.2 Timeliness and periodicity				X		Dissemination lags exceed GDDS recommendations, especially the printed monthly bulletin “Information Notes and Statistics” (NIS).	Renovation of the computerization of all the BCEAO’s accounting functions, reduction in the time taken to disseminate monetary statistics to 45 days.
	4.3 Consistency		X
	4.4 Revision policy and practice		X
5. Accessibility
	5.1 Data accessibility				X		Statistics are made available to selected users much before the public at large. No pre-announced release schedule. No comments are issued with the monetary statistics.	Publication of a preannounced release schedule.
	5.2 Metadata accessibility				X		No dissemination of methodological document to the public (apart from GDDS metadata in English on the IMF web site).	Dissemination to the public of the monetary statistics’ methodology.
	5.3 Assistance to users			X			In the NIS and on the BCEAO web site, no contact person is publicized; contact persons are mentioned in the GDDS metadata only.

0.1.1 The responsibility for collecting, processing, and disseminating statistics is clearly specified

The Central Bank of West African States (BCEAO) is responsible for preparing the external balance of payments (BOP) of the member states of the West African Economic and Monetary Union (WAEMU).

The authorization is provided in Article 1 of Annex III of Regulation R09/98/UEMOA of December 20, 1998 (the Regulation) concerning external financial relations of member states of WAEMU. The Regulation authorizes data collection from national and international agencies, either directly or through banks, financial institutions, the postal administration, and notaries (Article 8, Annex III of the Regulation). Each National Directorate of the BCEAO implements this obligation for enterprises and banks located in its national territory, whereas the BCEAO Headquarters carries out the obligation in respect of international organizations.¹⁹

By virtue of this authority, the BCEAO provides the technical secretariat for the Balance of Payments Committee, which is a national body for examining data collected and for validating balance of payments statistics. The Committee was created under Article 3 of the Regulation and is chaired by the minister in charge of finance for each member state of the WAEMU.

0.1.2 Data sharing and coordination among data producing agencies are adequate

In general, the provisions on data sharing and coordination among data producing agencies are adequate; their implementation, however, can sometimes prove limited. Data coordination takes place in two main frameworks:

• The Balance of Payments Committee, which is in charge of examining, validating, and authorizing the publication of the external accounts. All the government units that produce statistics for the balance of payments participate in this committee.

• The BCEAO, which is involved in many other forums of discussion and coordination of Burkina Faso’s macroeconomic statistical information, such as the Forecasting and Conjuncture Committee for Macroeconomic Management, the National Statistical Coordination Council, and the National Trade Balance Commission. Members of these committees discuss the data collected, the statistical techniques used, and the series produced.

In the specific context of merchandise trade, BOP committee members—the customs administration and the National Institute of Statistics and Demography (INSD)—provide the BCEAO with source data, deemed satisfactory. However, the reprocessing and adjustments the BCEAO performs for purposes of calculating the goods item in the balance of payments based on trade data are not always accompanied by an explanatory note on methodology, that may serve as formal and useful guidance to other interested agencies, in particular the INSD.

Furthermore, in view of the fact that Burkina Faso’s National Trade Balance Committee has not met since 1993, technical exchanges of views among statisticians on the specific subject of trade prove to be very limited.

0.1.3 Respondents’ data are to be kept confidential and used for statistical purposes only

The BCEAO is not allowed to disclose transactions or data of individual institutions. Article 2(1) of Annex III of the Regulation guarantees the confidentiality of information collected for purposes of compiling the balance of payments statistics. Only staff directly participating in compiling the balance of payments have access to these data, and they are forbidden to communicate them to any other person or agency not involved in the processing.

In correspondence sent to respondents, a reminder is included about the confidential nature of the data and the exclusive use of the information supplied for the sole purpose of compiling balance of payments statistics. These provisions are also indicated with each annual survey.

Article 2(2) of the Regulation prohibits the disclosure of individual data. This Article is supported by the provisions of the BCEAO Staff Regulations, as well as the by-laws, which provide penalties for staff members who use individual data for nonprofessional purposes. New staff regulations, adopted in February 2002, further strengthened the confidentiality expected of staff. The regulations forbid them in particular from making public statements during their tenure and after they leave (Article 4 of the Staff Regulations).

Other practical internal rules guarantee that the individual data cannot be used for other purposes than the compilation of statistics:

• Only BCEAO staff in charge of collecting and recording individual data have access to them. The other persons involved downstream, especially during the validation phase in the Balance of Payments Committee, have no access to individual data.

• After processing the data, the staff file and place them in the hands of the Unit Chief in charge of preparing the balance of payments.

• At the dissemination stage, only the aggregate data are presented in the detailed tables for external distribution. No other individual data are made public.

The physical security of the data is ensured on the premises of the BCEAO National Directorate. In addition, the computer systems prevent copying of files from any staff computer.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

Statistical reporting is mandatory under Article 1 of the Regulation. The Article establishes the BCEAO’s authority clearly and unambiguously to collect the data.

There are no reported or potential conflicts between the legal authorization to collect the necessary data for establishing balance of payments statistics and any other laws or regulations. Furthermore, the Balance of Payments Committee provides a forum for preventing any differences in approach in the implementation of the Regulation.

The obligation to report data is accompanied by a provision for penalty for failure to meet deadlines and or sending erroneous data (Regulation Article 10). The obligation is also dealt with under the exchange control violation legislation and the banking law. In practice, however, this legislation has never been invoked in Burkina Faso, as the BCEAO is currently promoting coordination and awareness on the part of respondents.

Within the assigned resources, the BCEAO takes due account of the workload of respondents and helps them to complete and submit the forms. Respondents have more than one month to return the completed forms. They are encouraged to contact the balance of payments correspondents at the BCEAO to clarify any ambiguity in the questionnaires or to obtain help when a complex operation has to be reported. Frequent contacts (by phone or by personal visits) by the BCEAO also help resolve the various problems that respondents face.

0.2.1 Staff, financial, and computing resources are commensurate with statistical programs

The number of staff is deemed insufficient to compile and disseminate balance of payments data, while the financial and computing resources are adequate.

The number of staff (four) of the Research Unit of the BCEAO National Directorate in Burkina Faso assigned to compiling balance of payments statistics are currently insufficient to deal with the tasks of collecting, preparing, and analyzing the data. What is more, two of them can, depending on the workload, be assigned to another section of the Research Unit. According to the workload standards set by BCEAO Headquarters, the Burkina Faso National Directorate is understaffed by two (one professional and one nonprofessional) for balance of payments statistics compilation. The Balance of Payments Section will be reinforced soon.

In general, the financial resources allocated to preparation of balance of payments statistics are sufficient in the context of the overall resources the BCEAO possesses. The budgetary resources are allocated within the framework of the statistical development program of the BCEAO management.

Adequate IT resources have been made available to the section preparing the balance of payments. Every officer has a computer and modern data processing technology allows for automatic processing of received data in the transaction-term (BPT) application of the balance of payments. The installed equipment is adequately distributed to ensure that data are efficiently processed and managed.

0.2.2 Measures to ensure efficient use of resources are implemented

At the level of the BCEAO National Directorate, there is no analytical identification of the cost of compilation of BOP statistics and the BCEAO fund planning covers only the budgetary and accounting management.

Regionally, the BCEAO has put in place internal procedures for the costs of the various tasks performed. It monitors efficiency in terms of the cost of activities for balance of payments statistics compared to other statistics and takes corrective action if there is any discrepancy. Both the BCEAO Headquarters and the National Directorate monitor cost-effectiveness.

0.3.1 Processes are in place to focus on quality

The BCEAO supports the principles of reliability and accuracy in balance of payments statistics. The internal arrangements include hiring and training highly qualified and knowledgeable staff, preparing surveys and other vehicles for collecting source data, moving from Balance of Payments Manual, fourth edition (BPM4) to Balance of Payments Manual, fifth edition (BPM5), and introducing a computer application to operate uniformly in all BCEAO National Directorates.

These measures are supplemented by internal and external controls at the BCEAO. Internal Control is performed by the Research Unit’s internal inspection section and the BCEAO’s internal auditor, who examines the enforcement of management rules and procedures in place in all units of the Bank. For example, the internal auditor of the Research Unit ²⁰ reviews the purpose and lists recipients of all questionnaires sent out, the response rates, and the preparation of internal notes by the units. External auditing essentially involves use of external assistants or experts, who can provide advice on organizational matters in the BCEAO.

0.3.2 Processes are in place to monitor the quality of the collection, processing, and dissemination of statistics

The Balance of Payments Committee addresses quality issues in the collection, processing, and dissemination of statistics: it influences the use of methodology by expressing opinions, and makes recommendations on the collection sources and strategies for improvements in data production.

Apart from the institutional users participating in the Balance of Payments Committee, the other users are not surveyed or questioned systematically on the quality of data. The BCEAO is made aware of their needs when they submit requests for information.

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

Quality enhancement is a permanent concern of the BCEAO authorities and is evidenced by their efforts to modernize balance of payments statistics compilation. However, some delays in the implementation of methods following the transition between BPM4 and BPM5 at the National Directorate and the consequent development of systems applications hamper further enhancements of quality.

The BCEAO National Directorate for Burkina Faso emphasizes the comprehensiveness of the source data collection. It also closely monitors publication deadlines within the automated application used for compiling annual data. However, following the adoption of the BPM5 during 1996–99, the BCEAO has not introduced yet the expected significant improvement to the compilation of BOP statistics. This involves chiefly the addition of supplementary information from the Bank for International Settlements and from the large enterprise balance sheet database (Centrale des Bilans). The computer application meant to compile quarterly balance of payments data in terms of settlement (BPR) is also on hold, although the BCEAO designed it already during 1996–99. Although the banks already report settlement based data, the IT application is not finalized. These delays hamper further improvements to the compilation methods for balance of payments statistics.

The BCEAO shares with its economic partners, chiefly the members of the Balance of Payments Committee, its most important options and considers the points of views of the members of the Balance of Payments Committee. However, it does not conduct surveys of external users nor does it invite them to make comments or express their wishes. The BCEAO is open to any requests for information on balance of payments statistics that external users may wish to submit, and it is always ready to provide them with any necessary explanations.

1.1.1 Statistics are compiled on an impartial basis

A shared decision-making process among the member states of the Union ensures the autonomy, objectivity, and independence of the BCEAO and, thus, the impartiality of the statistics. Because all the member states share in equal parts the BCEAO’s capital (Article 3 of the BCEAO Charter), all members adopt economic and monetary decisions, including those concerning balance of payments statistics. As the National Directorate depends in all matters (missions, resources) on BCEAO Headquarters, it enjoys some degree of independence in its statistical options.

Regarding the professionalism of staff, the BCEAO encourages its staff to undertake studies and attend regional or international training courses. Staff are also members of professional groups in Burkina Faso for exchanges of views on international transactions and balance of payments statistics.

The staff of the National Directorate are permanent employees and can only be dismissed for serious misconduct. Their recruitment is based either on competitive examination or on dossiers submitted to a Commission examining all applications. They are encouraged to develop their professional skills. They may produce memoranda or research papers, some of which are published in the Études et recherche (Studies and research) booklet of the Notes d’Informations et Statistiques (NIS). They are informed about training possibilities by BCEAO Headquarters and have access to training courses, seminars, colloquiums, and meetings with other experts.

1.1.2 Choices of sources and statistical techniques are informed solely by statistical considerations

All techniques of collecting and aggregating data meet statistical requirements, especially the standards recommended by BPM5. Although not published, internal methodological memos are generally intended to help the BCEAO balance of payments compilers in their tasks.

These memos transpose the BPM5 definitions, classifications, and methods without changing either the spirit or the letter of them.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The BCEAO has the right to comment on statistics publicly and to correct any errors of interpretation of the statistics it produces. For this purpose, its Research Unit monitors press and other media commentaries. In case of an erroneous commentary, the BCEAO sends corrective information to the editor of the publication. The media have thus far not erroneously commented on balance of payments data.

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The terms and conditions under which balance of payments statistics are compiled and disseminated are not systematically referred to in the official publications of the BCEAO. The publications should in particular draw the user’s attention to the responsibilities of the BCEAO and the ministry in charge of finance, in the collection, processing and dissemination of data ²¹ and on the accuracy of individual data communicated by respondents.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

Provisional aggregated balance of payments data are communicated to the Ministry of Finance and Budget, also a supplier of primary data used in the compilation, before they are submitted to the Balance of Payments Committee. This prior communication is indicated only in the GDDS metadata table, which describes the BCEAO’s activities; this table is currently available only in English on the IMF’s web site (http://dsbb.imf.org). Therefore, the BCEAO could easily refer to this information specifically for French language users in the documents it disseminates.

1.2.3 Products of statistical agencies/units are clearly identified as such

All the data disseminated by the BCEAO are done so under its authority and bearing its own insignia and identification. When users employ these data, they are asked to indicate their source. In the case of the balance of payments, the data are compiled and disseminated in the name of the Ministry of Finance and Budget.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

There is generally no advance notice of the introduction of a new survey or a new statistical technique. The BCEAO informs the users when new survey results or statistical techniques are introduced into the compilation of balance of payments data.

However, when the 1996 balance of payments was being prepared, respondents were notified in advance about the transition to BPM5. When the revised time series were disseminated, a detailed note explained the main methodological changes.

1.3.1 Guidelines for staff behavior are in place and are well known to the staff

Guidelines for staff behavior are clear and well known. At the time BCEAO staff are recruited, they undertake in writing to comply with professional secrecy and refrain from speaking publicly in the exercise of their functions. This provision thus covers the collection and processing of individual information.

After consulting the national authorities, the Governor of the BCEAO appoints the National Director for an unspecified period. The National Director’s tenure is revocable by decision of the Governor, again after consultation with the national authorities.

Newly arriving staff at the National Directorate, including trainees, are informed about the ethical practices and rules in force at the Bank. They are given a copy of the BCEAO Staff Regulations, as revised in February 2002, and the Bylaws of the Bank. Staff members are also regularly reminded of these rules.

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

The statistics are presented in accordance with the definitions, concepts, and classification of BPM5. For example, the concept of residence complies with the definition given in BPM5, both at the regional level of the Monetary Union area and in the national economies.

At the national level, all the items of the BPM5 standard components are presented, except for the long-term/short-term breakdown of other investments, only available for banks, and certain information that cannot be broken down at the collection stage, e.g., breakdown of other services. A number of other details are not available, such as the breakdown between business travel and personal travel, owing to insufficiently detailed sources.

At the regional level, the definitions and concepts of BPM5 are fully observed.

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices

In principle, Burkina Faso’s balance of payments covers all transactions between residents and nonresidents. The residence concept follows the BPM5’s recommendations and is based on the principle of the center of economic interest. Emigrants are considered to be nonresidents, and immigrants as residents. This treatment is clearly described in the technical notes prepared for use by the BCEAO’s balance of payments compilers.

The National Directorate conducts a specifically designed annual survey of the nonbank financial institutions and the West African Development Bank.

The international nonfinancial institutions, such as the WAEMU Commission in Ouagadougou, the Banking Commission in Abidjan, and the ASECNA Research and Development Institute in Dakar report their data directly to BCEAO Headquarters, which passes on the relevant information to the National Directorates.

The National Directorate and the BCEAO Headquarters work in close contact to prepare the regional balance of payments for the WAEMU countries. Using the detailed counterpart transactors identified in the national balance sheets, the BCEAO Headquarters consolidates the regional balance of payments. It nets out the transactions between residents of the countries forming the WAEMU, to obtain only transactions between residents of the Union and the rest of the world.

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices

At the national level, the institutional units as well as transactions are classified in conformity with BPM5. The same principle is applied at the regional level.

2.4.1 Market prices are used to value flows and stocks

Transactions are valued and recorded on the date of the transaction, at the effective exchange rate agreed to by the parties, in conformity with the principles set forth in BPM5. Grants in kind are recorded at their value reported at the time of their physical arrival. The data for other transactions are in general valued at their current market value on the date of the transaction. Freight and insurance data are calculated by balance of payments compilers based on customs data and are recorded on the date of submission of the customs documents.

For external public debt operations, the data for transactions are supplied by the Public Debt Directorate (DDP, in the General Directorate of the Treasury and Public Accounting of the Ministry of Finance and Budget). Expenditure and drawings under the debt are recorded based on the amounts disbursed.

Transactions reported in currencies are converted into CFA francs and valued based on the exchange rate of the transaction. Positions are valued on the last working day of the fiscal year; transactions derived from positions are converted using the annual average CFA franc/foreign currency rate.

External public debt data, supplied by the Debt Directorate, are transmitted using a valuation based on the exchange rate current on December 31 of the reference year. Capital and interest repayment operations are posted based on the exchange rate on the date of the transaction in question.

Transactions reported by enterprises and banks are transmitted in CFA francs to the national agency. The conversion rates used are those observed on the date of the transaction.

Regionally, the BCEAO Headquarters process data reported in currencies using the exchange rate on the date of the transaction for flows and the average annual rate for the valuation of stocks.

2.4.2 Recording is done on an accrual basis

Balance of payments transactions are recorded following the accrual principle except for goods (customs data) and income (cash basis), in compliance with the principles of the Balance of Payments Compilation Guide. These exceptions are clearly referred to in the BCEAO’s methodological memoranda to BCEAO staff.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices

Grossing/netting procedures comply with the principles established by BPM5. Transactions are recorded distinguishing between assets and liabilities. Current transactions are declared and recorded on a gross basis. Financial operations are declared and recorded separately by category, excluding fees and financial charges; each asset and liability component is broken down separately. Recording conventions are identical at both the national and regional levels.

3.1.1 Source data are collected from comprehensive data collection programs that take into account country-specific conditions

The overall data collection programs are primarily based on survey sources, while the source for bank payments is poorly developed and the coverage of private sector assets and liabilities is also weak. The primary sources of gross data include the surveys of enterprises, NGOs, some national institutions, banks (their own operations), customs data, and reports on the activities of some international institutions (United Nations Development Program (UNDP), the French Cooperation Administration). The National Directorate of the BCEAO periodically reviews all these sources to ensure that the data collected are complete. These data are mutually cross-checked to verify the information.

For purposes of the annual surveys that provide the main basis for compiling the balance of payments statistics, the National Directorate keeps the records from an array of different reporting entities. For instance, the surveys for 2003 are based on 312 questionnaires, of which 75 are for industrial and commercial enterprises, and 80 for NGOs. The Burkinabè press and Chamber of Commerce help detect changes of status of enterprises in the sample during the year in question.

Every two years, the National Directorate conducts a comprehensive program to revise data sources used for producing the balance of payments statistics. For example, to prepare the annual surveys of enterprises constituting the sample, the Directorate systematically revises its list of registered operators involved in Burkina Faso’s international trade. This step includes newly created enterprises and drops those that have stopped doing business.

Apart from the survey sources, the National Directorate collects biweekly information from banks (payments reports—CRP) on settlements by their customers with nonresidents, which, however, is not used in the published annual balance of payments statements. These data are collected and used wholly for preparing a new computer application, designed eventually to be the sole source of quarterly balance of payments statements. This source is currently undergoing reliability tests.²²

Further cross-checking possibilities are expected from two additional sources: (1) the data from the Centrale des Bilans (a database the BCEAO maintains that summarizes statements of corporate balance sheets) and (2) data from the Bank for International Settlements (BIS) on deposits by Burkinabè residents with nonresident banks. A strategy to incorporate these two sources is included in the table of plans for improvements in metadata of the General Data Dissemination System (GDDS)²³ for Burkina Faso. These sources could provide the quality control factors that will make it possible to verify and check the summary statements submitted by enterprises and banks.

In addition, the National Directorate receives from BCEAO Headquarters two other data statements:

• A statement of banks’ assets and liabilities, derived from the reports for prudential and monetary control purposes (positions statements DEC 2000). These aggregate data are used to calculate the financial operations account items, especially flows of other investments.

• A statistical statement on the gross and reserve assets and liabilities (External assets and liabilities of the Central Bank), which is used to record Burkina Faso’s reserves in the balance of payments statistics.

Finally, the external sector is constantly monitored, especially through press review. Privatizations of public enterprises are monitored to complete direct investment operations. Licenses granted to nonresident private operators of wireless telephony in 2000 were singled out for monitoring, because of the size of these operations and the identification of new operators.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

Source data on the whole approximate the definitions, classifications, valuation, and time of recording required. However, the scope of surveys is limited because of the low rate of response to surveys, which in a normal year averages 40 percent. This reflects largely the small number of replies from industrial and commercial enterprises and nongovernmental organizations (NGOs) in Burkina Faso. Nonetheless, the cross-checks that can be made thanks to the Directorate for Monitoring NGOs (Ministry of Economy and Development), help offset the inadequacy of the data on these two types of operators.

Several methods are used to correct customs data:

• Specific adjustments to certain categories of products, such as nonmonetary gold for imports;

• Warehouse movements under bond (difference between general and special trade);

• Value adjustments on the category of products, e.g., live animals exported, petroleum products imported, and mineral products—for example, the customs data are adjusted by the data from the Burkinabè Precious Metals Board (Comptoir Burkinabè des Métaux Precieux—CBMP) for gold exports; and lastly,

• The c.i.f./f.o.b. adjustment, which corresponds to the freight and insurance components of c.i.f. customs data.

The National Directorate of Burkina Faso evaluates informal trade based on movements of banknotes among bordering countries that are member states of the Union. All banknotes issued by the BCEAO are identical, except for an identifier on the front side of each note for the country in which it was initially issued.²⁴

According to the rule of management of banknotes that are legal tender in the BCEAO, each of the National Directorates regularly sorts circulating banknotes. When banknotes not belonging to the country of issue are identified, they are separated and returned to their country of issue.

The banknotes returned by countries bordering Burkina Faso at the end of this process make it possible to calculate the volume of fiat currency exported from the Burkinabè territory of issue. In the balance of payments, the amounts of such “external” banknotes are then broken down by national distribution key to duly account for undeclared current transactions. In Burkina Faso, these amounts are distributed among goods, services, and transfers,²⁵ and the country keys are revised every year.

The source data are collected in forms adjusted to the needs of drawing up the balance of payments. Thus, the Debt Directorate provides detailed data following the BPM5 presentation. The BCEAO has also developed some 40 other forms for the other categories of economic agents, of which a dozen or so are currently used.

Regionally, BCEAO Headquarters works from national data sent in by the National Directorate of Burkina Faso.

3.1.3 Source data are timely

In general, received source data are timely. When correspondence is sent out, respondents are told the deadline for returning the questionnaires they receive. In principle, questionnaires are expected back one month after being sent out. If a respondent fails to respond, reminders are sent out by mail. Depending on the size of the enterprise or the assumed volume of its foreign transactions, National Directorate personnel will visit the respondent one week after the reminder is sent out. Where the respondent is located in Bobo-Dioulasso, the staff of the Bobo-Dioulasso BCEAO branch takes part in the statistical data collection process.

3.2.1 Data compilation employs sound statistical techniques

In general, the data compilation employs sound statistical techniques. The gross data received from respondents are automatically processed after their manual capture by the staff of the National Directorate. Since 1996, a computer application developed by BCEAO Headquarters has been installed at all National Directorates. This application (Balance of Payments in Transaction Terms—BPT) performs automated verification, such as the logical comparison with the gross data submitted the previous year, or the adding up of subcomponents. The computer application is described in a single document accessible by the staff in charge of preparing the balance of payments. The Oracle database is used to store the captured data.

The survey forms are easy to complete and have been tested beforehand on a sample of respondents.

3.2.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

The other statistical procedures are based on appropriate methods for preparing balance of payments statistics. Examples are as follows:

• Merchandise trade: In conformity with the BCEAO Headquarters recommendation that the c.i.f/f.o.b. ratio be revised every year, the National Directorate complies, using an analysis of individual customs declarations.

• Informal trade: As indicated above (3.1.2), informal trade is evaluated based on the banknotes returned by the National Directorate of Burkina Faso.

• Transport services: Calculating revenues and expenditures for transport services is based on the average rates from a sample of import and export fares. The average rates are then applied to the amounts of imports and exports.

• Other investments: At the regional level, a correction is made in the other investments, money and deposits item, to neutralize external holdings in fiat money of another country in the CFA franc area. Thus, the assets and liabilities of a given country are on a net basis (after deduction of the assets of the other countries).

3.3.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide planning

The survey results are verified by comparing them against the results for the previous year. Because the whole population is targeted by the questionnaires, there is no problem with the choice of a sample.

However, limited corrective measures compensate the low response rate to the surveys mentioned above (3.1.2). The missing information is not fully accounted for in the results of surveys. This translates into inadequate coverage of subsequent data collection efforts.

Nonetheless, staff perform some cross-checking category by category. For example, they cross-check direct investment revenues in the current fiscal year against the stocks of direct investment in the previous year. The computer application automatically checks high-amount transactions at the capture stage, as well as after the analysis stage, when the data are inserted into the overall Burkinabè macroeconomic framework. Staff also confirm the data through direct contacts with respondents.

The data coming from banks on the transactions of their clients with nonresidents are not incorporated into the annual balance of payments statistical statement.

At the time the results are merged, the staff may reexamine basic data to correct any errors in the capture or classification. They may cross-check them against secondary sources. For example, they will cross-check acquisitions or disposals of securities declared by banks against information from the National Directorate’s Credit Unit. Given the annual periodicity of the balance of payments data, the source data are not revised, as the Directorate considers that all possible information is made available during the year’s compilation period.

The initial results consolidated by category of economic agents (known as “the centralized questionnaires”) are used to monitor the credibility of current data by comparison with previous years’ data.

3.4.1 Main intermediate data are validated against other information where applicable

The intermediate data derived from the surveys are matched against the data from administrative sources. Thus, the results from questionnaires to NGOs are matched against the data from the NGO Monitoring Bureau.

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated

Discrepancies are investigated. The centralized questionnaires are used to monitor major economic developments from one fiscal year to the next by category of economic agents. The behavior of certain series is regularly monitored by means of closely related series. For example, arrivals and departures of travelers are reconciled with the data from the Directorate General of Tourism and Hostelry of the Ministry in charge of Tourism.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

Errors and omissions are examined based on a ratio, which is calculated as the total amount of errors and omissions to the total amount of debits or credits (whichever higher). If this ratio (5 percent) is exceeded, verifications, adjustments, and cross-checks are initiated between current transactions and capital and financial operations and additional surveys, so as to reduce this rate. The addition of errors and omissions over several years should also be offset.

In accordance with the WAEMU method of preparing balance of payments data, the BCEAO Headquarters carries out the bilateral matching and comparison with the mirror data from other members of the monetary union. At this stage, any asymmetries are identified and reported to the National Directorates implicated in these bilateral discrepancies.

3.5.1 Studies and analyses of revisions are carried out routinely and used to inform statistical processes

Analyses of revisions are customarily conducted. The following serves to illustrate. The BCEAO produces provisional data, which it makes available to its partners in the official economic administration, mainly the Estimates Committee of the Ministry of Economy and Development. The data are also used in monetary planning at the national and regional levels of the BCEAO. These provisional data are not publicly distributed. Until the Balance of Payments Committee adopts them, they can be revised. The provisional data supplied as part of monetary planning are accompanied by a memorandum from the National Directorate explaining any changes from the initial estimates to the new estimates (“Review of the monetary program”).

However, the results of revision studies are based on the previously unavailable new data, and the methodology for preparing such statistics is not routinely revised.

4.1.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored

There is no regular mechanism for consulting users in the public. Their needs are considered only when they themselves make them known.

There is also no systematic consultation mechanism for institutional users. However, the institutional users, who are mostly members of the Balance of Payments Committee, frequently raise questions with the BCEAO, using the contacts they have established during the preparation of the balance of payments. These contacts allow for an exchange of views among institutions. For example, the INSD often submits requests for more details than the information disseminated.

In addition, the BCEAO is a member of economic forecasting groups: the National Debt Committee, the National Estimates and Conjuncture Committee, and the National Economic Policy Committee.

Although, a Statistics Coordination Committee of Burkina Faso is the decision-making authority on the methodological approach, the Committee in fact never meets.

4.2.1 Timeliness follows dissemination standards

The balance of payments data are disseminated about ten months after the end of the fiscal year, which is not in conformity with the GDDS recommendations (six to nine months).

4.2.2 Periodicity follows dissemination standards

Presently, periodicity meets GDDS recommendations on dissemination; balance of payments statistics in transaction terms are currently prepared and disseminated annually. Some components are prepared more frequently: the BCEAO publishes its net external assets and those of resident banks monthly in its Notes d’Informations Statistiques (NIS).

4.3.1 Statistics are consistent within the dataset

Dataset consistency is ensured. When major methodological changes occur (e.g., the transition from BPM4 to BPM5), the time series are systematically revised backward. Thus, in 1996, the balance of payments statistical series for the period 1986–95 were reorganized to follow the headings and presentation of BPM5, so as to ensure consistency of time-series.

The internal consistency of data is further supported by the level of detail of the series produced, which in Burkina Faso follows the standard classification set forth in BPM5. In the long term, errors and omissions present a stable profile, thanks to the controls described above.

Breaks in series are flagged for the users’ attention and are explained.

4.3.2 Statistics are consistent or reconcilable over a reasonable period of time

Consistent time series are available. Annual time series following the BPM5 format have been produced for 1986 onward. Any atypical developments are explained as appropriate in the context of the annual reports produced on this subject. Data prior to 1986 are still displayed according to the former Burkinabè presentation, which followed BPM4. Also, at the time of transition to BPM5, the major methodological changes were identified and explained to the users in notes that accompanied publication.

4.3.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks ²⁶

Balance of payments statistics are generally consistent with the other macroeconomic statistics. In fact, the BCEAO publishes reconciliation tables between customs data for merchandise trade and those for goods in the balance of payments. Consistency with the monetary data produced by the Research Unit’s Monetary Analyses Section is guaranteed by the fact that the balance of payments statistics are fed from the data obtained from the bank balance sheets of both the banks and the BCEAO.

Consistency with State budgetary and financial data is based on the same sources used to prepare the State flow of funds table (TOFE) and on the data sent by the Ministry of Finance and Budget to the National Directorate of the BCEAO.

As for the Balance of Payments Committee, which brings together all the institutional users of the balance of payments, the Committee members collegially examine overall consistency during the course of the fiscal year.

4.4.1 Revisions follow a regular, well-established, and transparent schedule

The final data for the year are not revised because the BCEAO believes that the total length of processing the annual data provides enough time to include all available data, and because it does not routinely introduce changes to the methodology.

However, the BCEAO revises preliminary balance of payments data disseminated to a limited number of users (see 3.5.1.); thus, the BCEAO produces at least two provisional data series before it arrives at the final data.

4.4.2 Preliminary data are clearly identified

The preliminary data are identified in the working documents sent to several policymakers by a heading “provisional,” included in all these documents and tables.

4.4.3 Studies and analyses of revisions are made public

Studies and analysis of revisions are not public, but only provided to Ministry of Finance users and members of the Balance of Payments Committee. In 1996, during the transition from BPM4 to BPM5, these respondents also received a memorandum explaining the change from the enterprise accounting procedures to the classification and definitions of BPM5. Users among the general public are not informed about these methodological memoranda (see 3.5.1.).

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

The balance of payments statistics have a uniform presentation, including analyses of events over the period, the international environment, and Burkina Faso’s macroeconomic context.

The balance of payments data are published once a year in a document entitled: Balance of Payments for Burkina Faso, Year N (“Balance des paiements du Burkina Faso, année N”), the most recent year available is 2001, under the imprint of the Balance of Payments Committee, Ministry of Economy and Budget of Burkina Faso. These data are also published in the Notes d’Information Statistiques (NIS), a quarterly series, and in the Rapport Annuel de la BCEAO, published by BCEAO Headquarters. Excerpts from the Information Note and Statistics (NIS) and the Rapport Annuel are posted on the BCEAO’s web site at www.bceao.int.

The quarterly NIS disseminate the tables, only with a comment on the balance of payments situation of the WAEMU zone as a whole, not with any country-specific commentary. Tables provide details and illustrate the analyses. The institutional users in the Balance of Payments Committee guide the presentation of tables in the papers.

5.1.2 Dissemination media and formats are adequate

Dissemination media and formats are adequate. The balance of payments data—both national and regional—are made available to the country’s main documentation and information centers. Although the formal adoption of data does not occasion any press releases, the Balance of Payments Committee edits a report—Rapport du Comité sur les Comptes Extérieurs du Burkina Faso, année N (Committee Report on the External Accounts of Burkina Faso, Year N)—for the Minister in charge of Finance and Budget. The Committee submits the report to the Council of Ministers and may also disseminate it.

Since March 2003, the Headquarters has posted data after adoption on the BCEAO web site (www.bceao.int), giving the complete series from 1986 to 2001.

5.1.3 Statistics are released on a preannounced schedule

Although BCEAO Headquarters sets a deadline for the release of data, the dissemination schedule is not preannounced. The rule is that data for a given year must in principle be released before the end of the following year. No further indication is provided. The annual balance of payments data are disseminated immediately after their adoption.

5.1.4 Statistics are made available to all users at the same time

Balance of Payments Committee members have access to the data before data are disseminated to the public. Immediately after the adoption of the data, these statistics are made available to all users and are delivered to a number of recipients, such as the university library and the IMF representation. The BCEAO makes no distinction among the various categories of users.

5.1.5 Nonpublished (but nonconfidential) subaggregates are made available upon request

Nonpublished aggregates may be sent to users on request. However, these data concern only balances already adopted any data that are not subject to any further revision.

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

No documentation on methodological aspects of data collection, production, or dissemination is made available to the public. Methodological memoranda compiled for the members of the Balance of Payments Committee are produced and may be communicated to external users who are aware of the existence of these memos and who ask the BCEAO explicitly for them. The BCEAO has not publicized the existence of these memos and does not disseminate them with the annual balance of payments results. Furthermore, the GDDS metadata for Burkina Faso are available on the IMF web site in English only, whereas the availability of metadata in French also would be highly desirable.

5.2.2 Levels of detail are adapted to the needs of the intended audience

The metadata are not prepared with the public in mind. There is no document designed to regularly inform the public.

5.3.1 Contact person for each subject field is publicized

No particular contact person is identified in the publications. If users have difficulty interpreting or understanding the content or methodology for compiling statistics, only the address and contact number of the BCEAO Headquarters is indicated in all statistical publications, such as the Notes d’Information Statistique and on the BCEAO web site: www.bceao.int.

In addition, the National Directorate’s Documentation Center is open to the public. However, this information is not indicated in the BCEAO publications.

5.3.2 Catalogs of publications, documents, and other services, including information on any charges, are widely available

At the regional level, there is a catalog of publications produced by BCEAO Headquarters, which is given out in the Notes d’Information Statistique, sent to all members of the zone. The BCEAO web site (www.bceao.int) also lists the published documents.

Table 5.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Balance of Payments Statistics

(Compiling Agency: BCEAO—National Directorate of Burkina Faso)

Table 5.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Balance of Payments Statistics

(Compiling Agency: BCEAO—National Directorate of Burkina Faso)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			Interagency cooperation weak between BCEAO and INSD.
	0.2 Resources			X			Balance of Payments Section is understaffed relative to needs for compilation.
	0.3 Quality awareness			X			Numerous qualitative improvements studied for years without being implemented.	ST—speed up introduction of improvements to enhance reliability and consolidate the statistics produced.
1. Integrity
	1.1 Professionalism			X
	1.2 Transparency				X		No methodological documents are made available to the public, some methodological changes are made without prior notice and the fact that ministries have access to information before it is disseminated is not made public.	ST—disseminate a methodological note to users.
	1.3 Ethical standards			X
2. Methodological soundness
	2.1 Concepts and definitions		X
	2.2 Scope		X
	2.3 Classification/sectorization		X
	2.4 Basis for recording		X
3. Accuracy and reliability
	3.1 Source data			X			Surveys, the main source for BOP data, have a weak response rate (40 percent).	Improve the response rate by monitoring each respondent and by broadening the survey base.
	3.2 Statistical techniques		X
	3.3 Assessment and validation of source data			X			Weak response rate to surveys is accepted, only partially compensated by secondary sources, but not taken into account by planning.
	3.4 Assessment and validation of intermediate data and statistical outputs		X
	3.5 Revision studies		X
4. Serviceability
	4.1 Relevance			X			Users’ needs are not taken into consideration.	Prepare dissemination formats and supports for the different audiences.
	4.2 Timeliness and periodicity			X			Timeliness of balance of payments aggregates does not follow the GDDS recommendations.
	4.3 Consistency		X
	4.4 Revision policy and practice			X			Revisions are only given to the institutional users and are not made public.	Disseminate revisions to broader public.
5. Accessibility
	5.1 Data accessibility		X					Prepare and disseminate press releases when balance of payments data are adopted.
	5.2 Metadata accessibility				X		Metadata are not disseminated and no distinction is made between different audiences.
	5.3 Assistance to users			X			The only contact for user assistance is the BCEAO headquarters in Dakar.

View Table

Table 5.

Burkina Faso: Data Quality Assessment Framework—Summary of Results for Balance of Payments Statistics

(Compiling Agency: BCEAO—National Directorate of Burkina Faso)

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
Element		NA	Assessment				Comments on Assessment	Plans for Improvement and Target Dates
			O	LO	LNO	NO
0. Prerequisites of quality
	0.1 Legal and institutional environment			X			Interagency cooperation weak between BCEAO and INSD.
	0.2 Resources			X			Balance of Payments Section is understaffed relative to needs for compilation.
	0.3 Quality awareness			X			Numerous qualitative improvements studied for years without being implemented.	ST—speed up introduction of improvements to enhance reliability and consolidate the statistics produced.
1. Integrity
	1.1 Professionalism			X
	1.2 Transparency				X		No methodological documents are made available to the public, some methodological changes are made without prior notice and the fact that ministries have access to information before it is disseminated is not made public.	ST—disseminate a methodological note to users.
	1.3 Ethical standards			X
2. Methodological soundness
	2.1 Concepts and definitions		X
	2.2 Scope		X
	2.3 Classification/sectorization		X
	2.4 Basis for recording		X
3. Accuracy and reliability
	3.1 Source data			X			Surveys, the main source for BOP data, have a weak response rate (40 percent).	Improve the response rate by monitoring each respondent and by broadening the survey base.
	3.2 Statistical techniques		X
	3.3 Assessment and validation of source data			X			Weak response rate to surveys is accepted, only partially compensated by secondary sources, but not taken into account by planning.
	3.4 Assessment and validation of intermediate data and statistical outputs		X
	3.5 Revision studies		X
4. Serviceability
	4.1 Relevance			X			Users’ needs are not taken into consideration.	Prepare dissemination formats and supports for the different audiences.
	4.2 Timeliness and periodicity			X			Timeliness of balance of payments aggregates does not follow the GDDS recommendations.
	4.3 Consistency		X
	4.4 Revision policy and practice			X			Revisions are only given to the institutional users and are not made public.	Disseminate revisions to broader public.
5. Accessibility
	5.1 Data accessibility		X					Prepare and disseminate press releases when balance of payments data are adopted.
	5.2 Metadata accessibility				X		Metadata are not disseminated and no distinction is made between different audiences.
	5.3 Assistance to users			X			The only contact for user assistance is the BCEAO headquarters in Dakar.