Botswana: Report on Observance of Standards and Codes (ROSC)—Data Module

This Report on the Observance of Standards and Codes data module provides a review of Botswana’s data dissemination practices against the IMF’s General Data Dissemination System, complemented by an in-depth assessment of the quality of the national accounts, consumer price index, producer price index, government finance, monetary, and balance-of-payments statistics. All three of Botswana’s statistics producing agencies have a legal and institutional framework that supports statistical quality and demonstrate an awareness of quality as the cornerstone of statistical work. All three agencies also demonstrates professionalism and provide ethical guidelines to their staff.


This Report on the Observance of Standards and Codes data module provides a review of Botswana’s data dissemination practices against the IMF’s General Data Dissemination System, complemented by an in-depth assessment of the quality of the national accounts, consumer price index, producer price index, government finance, monetary, and balance-of-payments statistics. All three of Botswana’s statistics producing agencies have a legal and institutional framework that supports statistical quality and demonstrate an awareness of quality as the cornerstone of statistical work. All three agencies also demonstrates professionalism and provide ethical guidelines to their staff.

I. Introduction

1. This Report on the Observance of Standards and Codes (ROSC) data module provides a review of Botswana’s data dissemination practices against the IMF’s General Data Dissemination System (GDDS). It is complemented by an in-depth assessment of the quality of the national accounts, consumer price index (CPI), producer price index (PPI), government finance, monetary, and balance of payments (BOP) statistics, using the Data Quality Assessment Framework (DQAF) developed by the IMF’s Statistics Department (STA). Socio demographic statistics are not covered in this assessment. This report was prepared by a mission from STA that visited Gaborone during October 11–24, 2001.1 The assessment is based on information gathered during the mission and information publicly available from hard copy publications and on Internet websites.

2. Section II includes an assessment of Botswana’s current data dissemination practices against the GDDS, and a review of its practices visáa-vis the Special Data Dissemination Standard (SDDS). Section III presents a summary assessment of six macroeconomic datasets, following dataset-specific DQAFs. Finally, Section IV provides a set of recommendations to improve Botswana’s macroeconomic statistics.

II. Data Dissemination Practice and the General Data Dissemination System: Current Dissemination Practices

3. Following its earlier commitment to participation in the GDDS, Botswana has designated Ms. Malebogo Kerekang from the Central Statistics Office (CSO) as the country coordinator. Preparation of metadata is the next step towards participation, which is publicly recognized through the dissemination of the country’s metadata and plans for improvement on the Fund’s Dissemination Standards Bulletin Board (DSBB). This process has not yet begun and the review of Botswana’s data dissemination practices against the dimensions of the GDDS is therefore limited in this section to the data dimension, i.e., the coverage, periodicity, and timeliness of Botswana’s core macroeconomic statistical frameworks and indicators (see Appendix 1, Detailed Assessments using the Data Quality Assessment Framework). Other dimensions of the GDDS are covered in the DQAF.

4. The main agencies producing official statistics in Botswana are the Ministry of Finance and Development Planning (MFDP); the Central Statistics Office (CSO), which is part of the MFDP; and the Bank of Botswana (BoB). The MFDP is responsible for the compilation of statistics on central government operations. The dissemination of statistics on central government operations is done by the MFDP, the CSO, and the BoB. The CSO is responsible for statistics related to the national accounts and prices. The BoB is responsible for the compilation of monetary, financial, and BOP statistics. Botswana provides access to these data through a number of publications and the following Internet websites:

5. In general, Botswana meets the GDDS specification for the core comprehensive frameworks and recommended indicators, with the following exceptions2:

  • There is no production index, which is a real sector data category. However, the mining index would be relevant given the share of this sector in total GDP.

  • Annual data on central government operations and central government debt—comprehensive frameworks of the fiscal sector—are disseminated with a lag of 11 to12 months compared to GDDS recommendations of six to nine months.

  • In the financial sector, coverage of the monetary survey is incomplete.

  • In the external sector, quarterly public and publicly guaranteed debt service schedules are available on request and monthly merchandise trade statistics are disseminated with a lag of 12 months, compared to the GDDS recommendations of one to two quarters and eight weeks to three months, respectively.

6. As Botswana has shown interest in subscribing to the SDDS, the current dissemination practices were also reviewed against SDDS requirements. The following points about the coverage, periodicity, and timeliness prescriptions of the data dimension serve to highlight some significant issues that would need to be dealt with prior to Botswana’s subscription to the SDDS.

  • Data on the industrial production index, data on general government operations, and the template on international reserves and foreign currency liquidity—are not currently compiled. External debt statistics are available on request.

  • In the real sector, quarterly national accounts (QNA) and the monthly CPI meet SDDS periodicity requirements. Data on the PPI, employment, unemployment, and wages/earnings are disseminated with a periodicity that does not meet the SDDS. The timeliness of the CPI exceeds the requirements of the SDDS, but the timeliness of all other data categories falls short of the Standard. Botswana’s QNA is at a formative stage and is expected to become a standard statistical product—disseminated with a lag of one quarter after the end of the reference period—by the end of the national accounts accounting year 2001/02.

  • In the fiscal sector, the periodicity and timeliness of data on central government operations and central government debt do not meet the SDDS requirements. Monthly data on central government operations are compiled but not disseminated.

  • The data categories included in the financial sector meet the periodicity but not the timeliness prescribed by the SDDS.

  • In the external sector, the periodicity of all data categories, except the BOP, are in accord with SDDS requirements. Compilation of quarterly BOP has commenced with improvements being addressed prior to dissemination. While timeliness of the international investment position exceeds SDDS requirements, timeliness of data on merchandise trade and international reserves are not in accord with the Standard.

III. Summary Assessments of Data Quality

7. Interest in assessing the quality of data derives from the objectives of complementing the SDDS and the GDDS with a consideration of the quality of the data being disseminated and of focusing more closely on the quality of the data that underpin surveillance of countries’ economic policies. Against this background, STA has developed a tool to provide a structure and a common language to assess data quality. The DQAF comprises a generic framework and a set of dataset-specific frameworks. The frameworks cover five dimensions of data quality—integrity, methodological soundness, accuracy and reliability, serviceability, and accessibility—and a set of prerequisites.3

8. Botswana’s macroeconomic statistics and statistical base are broadly adequate for surveillance purposes. Nevertheless, IMF staff identified shortcomings in some statistical practices that could hinder the accurate and timely analysis of economic and financial developments and the formulation of appropriate policies. The main findings are detailed below.

A. Prerequisites of quality

This category in the DQAF identifies conditions within the agency in charge of producing statistics that have an impact on data quality. The elements within the category refer to the legal and institutional environment, resources, and quality awareness.

9. The Central Statistics Office (CSO) has the legal obligation, under the Statistics Act of 1967, to produce the national accounts, the CPI, and the PPI. However, the mechanisms for data sharing between the CSO and other agencies and for promoting voluntary statistical reporting by survey respondents leave room for improvement.

10. Budgetary and computing resources are commensurate with the statistical needs for the compilation of most real sector statistics. Problems with recruiting and retaining professional staff have an adverse effect on development work. Available staff resources are constrained, and developmental work may stretch existing staff resources. Efficiency in the use of resources could be augmented by a review of professional staff duties and operational procedures, and professional training in national accounts methodology is needed. However, it should be noted that a Population Census, which requires the allocation of significant resources—including those usually assigned to the compilation of macroeconomic statistics—was undertaken in Botswana in 2001.

11. While the CSO’s mission statement reflects quality awareness, established processes for monitoring and addressing emerging data requirements are not utilized. Against the background of an expanding financial sector and the assignment of credit ratings to Botswana by international rating agencies, such processes will assume increasing importance. In general, management of the CSO is quality-oriented. Although staff know that the PPI is of poor quality, no corrective action has been taken due to its low priority and the lack of resources.

12. The Ministry of Finance and Development Planning (MFDP) collects, processes, and disseminates central government statistics in compliance with the Finance and Audit Act (1970). The fiscal and financial information disseminated by the MFDP is compiled by the Accountant General’s Department using the computerized accounts system. The MFDP’s cooperation with other agencies—particularly the CSO and the BoB that publish GFS—is good, and there is no duplication of effort. No consolidated final data are compiled on the local government subsector, and no single entity has been assigned these responsibilities. However, final data are published in audited reports for each local government unit.

13. The staff resources dedicated to compiling GFS are adequate, and efforts are being made to encourage the retention of experienced staff through short courses and faster promotions. The accounting and GFS systems are not computerized and are not adequate for accurate and timely reporting. However, initial steps have been taken to computerize these systems. Processes are not fully in place to monitor and improve the compilation and dissemination of GFS. Trade-offs among dimensions of quality are not acknowledged.

14. The Bank of Botswana Act, 1996 and the Banking Act, 1995 provide the Bank of Botswana (BoB) with the legal framework for the collection and dissemination of financial statistics. The BoB also has responsibility for the collection and compilation of data on external transactions of individuals, but the BoB is not designated as the compiling agency for the BOP. In accordance with the BoB Act, responding to statistical surveys conducted by the BoB is obligatory, and the confidentiality of data is guaranteed. The Law also includes provision for penalties in cases of noncompliance. To date, this provision has not been used as all respondents comply with the reporting requirements.

15. Although BoB’s managers are aware of the importance of quality in building trust among users of monetary statistics, there are no systematic processes in place to promote quality control and improvement of the statistics. The BoB does not resort to an internal or external advisory body to review the quality of its statistics nor seek regular feedback from users regarding timeliness, accuracy, and comprehensiveness.

16. Data collection and processing for BOP are supported by a fully integrated purpose-built computer system. Although the BoB staff appears competent, more internal and external training in statistical methodology is desirable. Staff and computing resources are adequate.

17. It should be noted that the effects of HIV/AIDS would continue to have significant negative impact on human resources in Botswana and affect all statistics producing agencies.

B. Integrity

Integrity identifies features that support firm adherence to objectivity in the collection, compilation, and dissemination of statistics so as to maintain users’ confidence. Elements refer to the professionalism and ethical standards that should guide policies and practices, which should be reinforced by their transparency.

18. Professionalism and ethical standards in statistical policies and practices underpin the compilation of real sector statistics by the CSO. However, the transparency of statistical practices and policies should be enhanced by disseminating information on major revisions to data, publicizing instances of prior internal government access to statistics, and identifying the agency and unit that compile the data on the Internet websites. The new Statistics Act is expected to make the CSO a more autonomous agency. The CSO gives advance notification of some major changes in methodology, source data, and statistical techniques to some users, but this is not comprehensive. The CSO compiles statistics on an impartial basis, and choices of sources and methods are based on statistical considerations. The organization scrutinizes all media comments on their statistics and follows up on any misrepresentation.

19. The Finance and Audit Act, and Financial Instructions and Procedures call for the MFDP’s proper use of data sources in compiling GFS. While professional and ethical standards are adequate, some aspects of transparency could be enhanced, particularly those concerning restrictions on public access to current year GFS and those pertaining to the advance notification of major changes in methodology, source data, and statistical techniques. Other aspects of transparency, such as clear identification of statistical products, breaks in the time series, and approval processes for publication of GFS, appear satisfactory.

20. Professionalism, impartial statistical techniques, and ethical standards support the integrity of the compilation of monetary and BOP statistics by the BoB. The terms and conditions of the compilation of these statistics are known to the survey respondents and other data providers; these terms and conditions are also accessible to data users in hard copy and electronic format.

21. The BoB environment fosters the integrity of the statistics. Since the basic source of information for monetary statistics is the balance sheet of the depository corporation, there is little room to influence the collection process. The public can have access to the basic terms and conditions under which monetary statistics are compiled and disseminated, as the BoB posts the BoB Act and the Banking Act on its Internet website.

22. Though not publicized, there is a practice of internal government access to the BOP data before their release for the Budget Speech and the central government’s Annual Economic Report. The choice of sources and methods is governed by statistical considerations. The independence of the BoB in the compilation of BOP statistics strengthens confidence in the integrity of data.

C. Methodological soundness

Methodological soundness refers to the application of international standards, guidelines, and agreed practices. Application of such standards, which are specific to the dataset, is indicative of the soundness of the data and fosters international comparability. Elements refer to the basic building blocks of concepts and definitions, scope, classification and sectorization, and basis for recording.

23. Botswana’s annual national accounts are based primarily on the concepts and definitions of the 1968 System of National Accounts (1968 SNA). Efforts at alignment with the 1993 SNA are being undertaken within the context of the United Nations (UN) milestone approach adopted by the Southern African Development Community (SADC), and some adaptations including the classification of mineral exploration as gross fixed capital formation, have been completed. The delimitation of the economy and the production and asset boundaries are in accordance with the 1968 SNA. Application of Botswana Standard Industrial Classification 3 (BSIC Rev. 3)—a national adaptation of the International Standard Industrial Classification of All Economic Activities Revision 3 (ISIC Rev. 3)—is underway, but other major classifications (e.g., Classification of Individual Consumption by Purpose—COICOP—and Classification of the Functions of the Government—COFOG) are not utilized. Production and capital formation accounts exist for each industry; income and outlay accounts and capital-finance accounts cover a full set of institutional sectors. Government expenditures are recorded on a cash basis, and the output/inventory valuation of diamonds are based on the export prices as a proxy for market prices.

24. The CPI methodology is reasonably consistent with internationally endorsed standards. The introduction of Classification of Individual Consumption by Purpose (COICOP) is planned for the rebasing expected in 2004. The weights for the CPI are out-of-date and will not be updated for a few years. Prices are collected in urban and some readily accessible rural areas. The CPI measures the changes in price of a representative basket of goods and services purchased by all types of households; prices are actual transaction prices paid by consumers.

25. The scope of the PPI covers all domestic manufacturing activities, but the mining and quarrying industry is not included. The weights are based on output but do not include own account capital formation. The prices collected represent transactions at producer prices.

26. Concepts, definitions, scope, and sectorization in the GFS are broadly consistent with the 1986 GFS Manual. The classification systems for central government are mostly in line with internationally accepted guidelines, with a few important exceptions. For example, “mineral revenue,” which consists of royalties and dividends and of mineral tax, is classified as tax revenue; royalties and dividends should be classified as nontax revenue, according to international standards. All transactions are valued at current market prices and are recorded on a cash and gross bases. Even refunds of revenue are recorded on a gross basis (the 1968 GFS Manual recommends net recording in this case). Transactions in foreign currency are converted to pula using the exchange rate for the day on which the transaction takes place. Debt stocks are valued at face value, and those denominated in foreign currency are converted to pula using the midpoint market exchange rate at the end of the reporting period.

27. For monetary statistics, the Monetary Statistics Section (MSS) prepares a monthly monetary survey that aggregates the balance sheet of the BoB and the consolidated balance sheet of the five commercial banks operating in the country. These commercial banks account for 95 percent of the deposits and 80 percent of the lending of the banking system. Not included in the monetary survey are four other deposit-taking institutions (two merchant banks, one building society, and one savings bank). To compile the survey of the central bank and commercial banks, the MSS mostly follows the methodology recommended in the Fund’s MFSM.

28. The principles underlying the classification of financial instruments and the sectorization of institutional units are broadly consistent with the MFSM. However, commercial banks mistakenly classify their accounts between residents and nonresidents mainly according to the nationality of the account holder, rather than the principle of “center of economic interest.” Both the BoB and commercial banks apply market price valuation of financial assets and liabilities, and they record assets and liabilities on both an accrual and gross bases.

29. Botswana’s BOP statistics are compiled in conformity with the conceptual framework of the fifth edition of the IMF’s Balance of Payments Manual (BPM5). The definitions of residents are in conformity with BPM5 concepts of economic territory, residence, and center of economic interest. However, commercial banks define nonresidents by nationality, as noted above. Except for the realization of capital gains on reserves, transactions are classified and sectorized in accordance with BPM5, and no transactions with nonresidents are systematically excluded. As a general principle, transactions are valued at market prices, and when transactions are derived from stock data, exchange rate changes and revaluations are excluded, whenever possible. Investment income is recorded on an accrual basis, and grossing and netting procedures are consistent with internationally accepted guidelines.

D. Accuracy and reliability

Accuracy and reliability identifies features that contribute to the goal that data portray reality. Elements refer to identified features of the source data, statistical techniques, and supporting assessments and validation.

30. The source data for the annual national accounts are derived from the comprehensive programs based on surveys and administrative records and complemented by statistical outputs from within and outside the CSO. However, shortcomings exist in the coverage of data on small-scale service sector enterprises, the informal sector, and the operations of unregistered construction companies. The sample frame for annual and quarterly business surveys relies on an up-to-date business register. The 1993/94 Household Income and Expenditure Survey (HIES) and the 1993/94 Social Accounting Matrix (SAM) provide the bases for benchmarks used in Gross Domestic Product (GDP) estimates, including the marketed component of household final consumption.

31. Differences in accounting periods and in the timing of recording between the national accounts and data sourced from the BOP and government finance adversely affect accuracy. The timeliness of source data are affected by low survey responses that warrant follow-up and by delays in the compilation of trade statistics. Adjustments to intermediate consumption for holding gains/losses on inventories are not made. Information on sampling and nonsampling errors, and on imputation methods for nonresponses for business surveys are not available. Ad hoc assessments of potential discrepancies of major intermediate data are done, but checks mainly occur only after discrepancies are identified. Independent estimates of GDP are not done. Thus, the size of the statistical discrepancy, which is published in terms of percentages of GDP, cannot be taken as an indicator of accuracy. The magnitude of revisions to GDP estimates are observed and reported, but studies and analyses of revisions are not undertaken.

32. The CPI is compiled from generally sound source data and methods. However, collected prices for house rentals are not representative; owner-occupied dwelling services are not included; and temporarily missing price quotes use the last reported price. The weights and prices are obtained from the HIES data collection program, but the survey is not run frequently enough. Revision studies are not applicable as data are assumed final when first released.

33. The PPI samples are too small to be representative of the survey universe. Two different samples produce quite different results. The weights and price quotations do not cover enterprises with less than five employees or household producers. Discontinued products are replaced with similar ones. Revision studies are not applicable as data are final when first released.

34. In GFS, annual data on budgetary central government and each unit of local governments are reported separately. The data cover the full range of financial stocks and flows, but with a 12-month lag for central government and up to a two- to three-year lag for local governments. These lags are due to lengthy reporting and auditing processes favoring accuracy rather than timeliness. The availability of data of these subsectors would allow the consolidation of data into a general government sector. There are no current plans to collect data for a consolidated local government subsector. Only small discrepancies exist between the MFDP’s, the CSO’s, and the BoB’s fiscal data since the MFDP is the compiler of GFS.

35. Apart from comprehensive annual data, the MFDP produces detailed monthly central government data for internal policy purposes with a lag of three to four weeks. No revisions studies are made as the GFS are based on final audited information.

36.Monetary statistics are compiled from the balance sheets of the BoB and the five commercial banks and from 21 schedules submitted by commercial banks for supervisory and statistical purposes. Commercial banks report their monthly data within the 21 days mandated by the Banking Act. However, the Accounting Department of the BoB submits its balance sheet to the MSS with a lag of more than 10 weeks. Consistency checks are performed manually by the MSS, but there is no routine cross-checking of data with other sources of information nor any current computer programs permitting comprehensive data collection.

37. The BOP is based on surveys of the private sector and supplemented with source data derived from an International Transactions Reporting System (ITRS), which was retained following the relaxation of exchange controls. BOP data are presumed to be accurate due to the high rate of response from enterprises (90 percent in the quarterly surveys and approximately 50 percent in the annual survey). The quarterly survey covers the largest enterprises only, whose foreign liabilities account for about 90 percent of total foreign liabilities. Verification procedures include consistency checks within survey forms, reconciliation of flow data with stock data, and built-in checks in support of the quantitative analysis of data. Effective scrutiny of individual responses from the surveys contributes to overall accuracy and reliability. Even though the survey-based approach is run in parallel with the ITRS, cross-checking is not done. While most of the data sources are timely, merchandise trade data from the CSO are not. Data revisions are assessed only occasionally, and some of the methods for estimating freight on imports and travel need to be updated. Net errors and omissions are monitored at the end of each compilation period. A bilateral comparison/reconciliation of the financial account was conducted with South Africa’s data in 2001.

E. Serviceability

Serviceability focuses on practical aspects of how well a dataset meet users’ needs. Elements refer to the extent to which data are relevant, produced and disseminated in a timely fashion with appropriate periodicity, are consistent internally and with other datasets, and follow a predictable revisions policy.

38. The national accounts program responds systematically only to the statistical needs of government while the needs of the private sector are monitored on an ad hoc basis. Periodicity and timeliness of the annual GDP estimates follow GDDS recommendations, and the CSO has commenced compilation of the QNA and is aiming for a three-month dissemination lag by mid-2002. Consistent time series data are available, and efforts are made to adjust historical data to account for methodological developments and also to avoid discontinuities arising from changes in data sources. Revisions follow a regular, though not well publicized procedure, and provisional and final data are clearly identified. The lack of a unified accounting period for the major statistical frameworks precludes reconciliation among data sets.

39. No systematic approach is used to assess the needs of users for either the CPI or the PPI. The CPI is published the second week after the reference month. Dissemination of the PPI takes place within two months of the reference month. The CPI is produced monthly, but the PPI is produced every third month. The CPI is consistent over time, but the PPI is not. Because its poor quality is recognized, no checks are made to ensure PPI consistency with other indices. The CPI is only revised in exceptional circumstances to correct errors, and the PPI has never been revised; these practices are known to users. However, users are not informed of the impact of revised weights on the CPI time series.

40. Although GFS on consolidated central government are compiled on a monthly basis, the first data released to the public are quarterly, compiled by the MFDP and disseminated by the BoB with a two- to three-month lag. In addition, annual data are disseminated by the MFDP with a 12-month lag. The annual financing data published by the MFDP are broadly consistent with monetary data. Annual data on local government operations, in a consolidated form, are not published, but data on the operations of each local government agency are published with a lag of two to three years.

41. The MSS introduces methodological changes to the monetary statistics when deemed relevant. The monetary survey is disseminated on a monthly basis, but since the BoB’s balance sheet is available to the MSS ten weeks after the reference month, the monetary survey is usually compiled with a three-month lag. Monetary statistics are released to the public only when the Botswana Financial Statistics (BFS) is published, which further impairs the timeliness of the information. Published data are usually final, but any necessary revisions are identified in the introductory notes of the BFS.

42. Botswana compiles and disseminates annual BOP data within two months after the end of the year, thereby meeting the timeliness and periodicity of the GDDS. The data are disseminated on a preliminary basis in the BoB’s Annual Report and the monthly BFS. Estimations are made for merchandise trade and other items based on incomplete information; revisions do not follow a publicized procedure, but new source data are incorporated as soon as they become available. The Annual Report contains a survey of BOP developments while the data are provided without comments in the BFS. Data published cover an 11-year period. BOP data appear to serve the needs of the user community as regular enquiries about them show that the data are being utilized. Data are consistent with monetary data and external debt but not with national accounts. The lack of a unified reporting period for the major statistical frameworks precludes reconciliation among dat sets. Revisions, as a result of methodological changes, are explained at the time of data publication, and occasionally the nature of revisions is investigated, and data sources and compilation procedures reviewed.

F. Accessibility

Accessibility deals with the availability of information to users. Elements refer to the extent to which data and metadata are clear and easily available and to which assistance to the user is adequate to help them find and use the data.

43. Dissemination formats for annual national accounts statistics are adequate. They include hard copy publications and electronic dissemination on the CSO’s Internet website. Provisional estimates are first released in the MFDP’s publication Annual Economic Report (which coincides with the Budget Speech presentation in the National Assembly) in February, but are only disseminated in the following quarterly publication of the CSO’s Statistical Bulletin. While the announcement of the date of the Budget Speech approximates to a preannounced release date for the national accounts, no formal advance release calendar exists. The CSO publishes an annual hard copy wall calendar that identifies the expected statistical outputs for each month, but release dates are not identified. Comprehensive and up-to-date metadata are not available. A contact person for the national accounts is identified in hard copy publications, and a catalogue of publications is widely available.

44. There is no advance release calendar for the CPI or PPI. Data are first released as a CSO’s Stats Brief with commentary and charts, before appearing in more formal publications and on the CSO’s Internet website. Metadata are available for the CPI, but not for the PPI. Published CPI data are broken down into 12 groups; additional nonconfidential breakdowns of the CPI are supplied on request; for the PPI, all such data are considered confidential. Contact persons are not publicized for either index, but most users know who to call.

45. The annual GFS on central government in the MFDP’s Annual Statements of Accounts allow major aggregates and balancing items to be identified and related to detailed underlying data. These MFDP’s accounts provide information with similar detail to that set out in 1986 GFS Manual tables. Both the CSO and the BoB publish central government data in such a way that they facilitate proper interpretation. Quarterly GFS, in summary presentation, are published in the BoB’s BFS. In the MLG’s Recurrent Budget Publication of Local Governments, which includes all fiscal operations of local government agencies, partial actual data on local governments are presented. The release date of official GFS publications is not preannounced, but all statistical publications are made available to all users simultaneously. Documentation on concepts, definitions, classifications, and methods is insufficient, e.g., differences from internationally accepted standards are not identified. No contact person appears in annual GFS publications or on the MFDP’s Internet website, but prompt and knowledgeable support service is made available to users.

46. Monetary statistics are accessible through the BFS, which is also posted on the BoB’s Internet website. The internal deadline for publication of BFS is not always met, and the public is not aware of this deadline. Nonpublished but nonconfidential data are available upon request. In the BFS, there is a list of the publications of the BoB, with their corresponding prices. There is not enough documentation on concepts, scopes, classifications, basis of recording, data sources, and statistical techniques.

47. Dissemination formats for BOP data and metadata include electronic dissemination, the Budget Speech, MFDP’s Annual Economic Report, press release together with the BoB’s Annual Report, and monthly and annual hard copy publications of the BoB and the CSO. The Budget Speech presents only aggregated BOP, but further releases are presented in a way that facilitates data analysis. However, the financial accounts lack clarity, primarily because no distinction is made between changes in foreign assets and liabilities. While the announcement of the date of the Budget Speech approximates to a preannounced release date for the BOP, no formal advance release calendar exists. BOP data are not released simultaneously to all parties. The analytic presentation of data is broadly in accordance with the BPM5, but the presentation of the financial account lacks clarity primarily because no distinction is made between foreign direct investment in Botswana and abroad, changes in portfolio investment assets and liabilities, and changes in other foreign assets and liabilities. Information about the detailed methodology and sources is available internally in a procedures manual, but it is not published and not available on request. Methodological notes are published in the IMF’s Balance of Payments Statistics Yearbook. Nonpublished but nonconfidential data are available upon request. A contact person is not publicized, but the BoB can be contacted in writing or by phone.

IV. Staff’s Recommendations

48. Based on the results of the data quality assessments, technical discussions with the Botswana authorities in the respective statistics-compiling agencies, and responses from data users, the following measures are proposed to (a) enhance the organizational structure of the statistical system, (b) bring Botswana’s statistical system in line with international standards, and (c) improve the usefulness of the data. These recommendations, which are subdivided into short- (less than a year), medium- (one to three years), and long-term (more than three years), build on the authorities’ plans as shown in Table 1. In addition, the most critical measures to address deficiencies in respect to these points and make economic and financial statistics more adequate for surveillance are identified as high priority.4

General Recommendations


  • As a priority, prepare GDDS metadata on (a) current statistical production and dissemination practices and (b) plans for short- and long-term improvements covering the data dimension (coverage, periodicity, and timeliness), the quality, the integrity, and the access by the public, for all sectors.

  • As a priority, strengthen data sharing and coordination among major agencies, especially with a view to reconciling major data sets with different accounting periods. In addition, data sharing and coordination would help to address key statistical issues including the basis for recording diamonds and emerging source-data needs (particularly monthly merchandise trade statistics) for the compilation of timely quarterly national accounts and quarterly balance of payments.

  • Make available more data and metadata for all sectors premised on the need for rapid and wide access. For example, disseminate monetary data through media such as press releases and on the BoB’s Internet website, in addition to the monthly BFS.

  • Develop and disseminate advance release calendars for all sectors.

  • Enhance transparency of statistical policies and practices by, e.g., publicizing practices relating to internal access to data prior to release and providing advance notice of major changes in source data and compilation methods.

  • Identify a specific contact person in hard copy publications and on Internet websites.

Medium term

  • Promote training in statistical methodologies, including through participation in the IMF Institute’s methodological courses, as well as the regular training of all personnel.

  • Regarding survey-based collection systems, disseminate information on data biases, response rates, sampling techniques, sampling, and nonsampling errors, and imputation methods for nonresponses.

  • Publish articles (or republish existing articles in convenient form) dealing with the methodology used in compiling macroeconomic statistics.

National Accounts


  • As a priority, building on the regional SADC approach, advance the work program for implementing 1993 SNA with an increased focus on conceptual adherence. Initially the implementation could be limited to the GDP estimates.


  • Develop proactive methods for encouraging voluntary response to survey-based statistical enquiries and mechanisms for data-user feedback.

  • Expand the work program for implementing 1993 SNA to include the development of financial accounts and balance sheets.

Consumer Price Index and Producer Price Index


  • As a priority, redesign the PPI to bring it up to international standards as regards the sample size and validation procedures.

  • Compile and disseminate the PPI on a monthly basis.


  • Implement a strategy for collecting price observations on house rentals reflective of existing market conditions and integrate same into the CPI.

  • Extend the PPI to include the mining and quarrying industry.


  • Following the next change of the CPI weights to 2002/03 (expected in 2004), regularly update the weights every five years, at least.

Government Finance Statistics


  • As a priority, disseminate monthly budgetary central government statistics within one month after the end of the reference period and central government debt data within one quarter after the end of the reference period.

  • Develop a migration plan for implementation of 2001 GFS Manual.


  • Emphasize the automation in the compilation of GFS while introducing the computerization of the accounting and budget system.

  • Shorten the current two- to three-year lag in dissemination of data on local governments.

  • Compile consolidated annual GFS on the general government sector (consolidated central and local governments) and provide time series harmonized initially with 1986 GFS Manual methodology and later with 2001 GFS Manual methodology.

  • Design and implement a phased work program to improve documentation on the concepts, definitions, classifications, basis of recording, and data sources used to compile GFS.

Monetary Statistics


  • As a priority, arrange for the electronic transmission of the BoB’s monthly balance sheet to the Monetary Statistics Section as soon as it is processed rather than awaiting audit.

  • As a priority, instruct the commercial banks to classify their accounts into resident and nonresidents accounts according to the principle of “center of economic interest” rather than according to the nationality of the accounts holder, as is the case at present.

  • Request the electronic submission to the BoB of the balance sheets and the accompanying schedules of the commercial banks and other deposit-taking institutions.

  • Routinely cross-check the monetary data with information produced by other sections or departments of the BoB and by other official agencies.


  • Install and make operative software that would allow automatic consistency checks and flagging, seasonal adjustment of series, and the building of a comprehensive monetary database.

Balance of Payments


  • Improve the presentation of the financial account by showing separately (a) foreign direct investment in Botswana and abroad, and transactions in foreign and domestic securities; (b) changes in foreign assets and foreign liabilities of the monetary authorities, government, banks, and other sectors; and (c) arrears as part of exceptional financing in the analytic presentation.

  • Strengthen selected classifications and methods: (a) Reclassify realized capital gains on reserves; (b) support the resident-nonresident classification of banking source data according to the principle of “center of economic interest”; (c) eliminate transactions between residents in the transportation account; and (d) update some estimation methods for freight payments on imports and travel.


  • As a priority, accelerate the development of quarterly balance of payments statistics. An essential feature of this effort would be to put in place timely compilation of monthly merchandise trade statistics.

Table 1.

Botswana—Data Quality Assessment Framework: Summary Presentation of Results

article image
article image
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

The mission team was led by Mr. Gérard G. Raymond and comprised of Messrs. Paul Austin, Johann Bjorgvinsson, Jose Maria Cartas, and Henri Hoezoo (all STA), and David Hughes (Expert).


An assessment of the encouraged data categories under the GDDS is also found in Appendix 1, Detailed Assessments using the Data Quality Assessment Framework.


See also the Generic Framework set out in Appendix III, Detailed Assessments using the Data Quality Assessment Framework.


This ROSC data module, including the staff’s recommendations, provides a “snapshot” assessment of statistics in Botswana as of October 2001. Actions already taken or announced to be taken by the authorities since October 2001 are included in the separate document “Response by the Authorities.”