The Report on the Observance of Standards and Codes (ROSC)—Data Module provides an assessment of Botswana’s macroeconomic statistics against the recommendations of the General Data Dissemination System, complemented by an assessment of data quality based on the IMF’s Data Quality Assessment Framework, July 2003. It reveals that there has been a strengthening of the relevance of the data, the transparency of statistical policies and practices, and the accessibility of data and metadata. Staff of statistical agencies also exhibits a high level of professionalism and ethical standards.

Abstract

The Report on the Observance of Standards and Codes (ROSC)—Data Module provides an assessment of Botswana’s macroeconomic statistics against the recommendations of the General Data Dissemination System, complemented by an assessment of data quality based on the IMF’s Data Quality Assessment Framework, July 2003. It reveals that there has been a strengthening of the relevance of the data, the transparency of statistical policies and practices, and the accessibility of data and metadata. Staff of statistical agencies also exhibits a high level of professionalism and ethical standards.

I. Overall Assessment

1. Botswana began participating in the General Data Dissemination System (GDDS) and started posting its metadata (including plans for improvement) on the Dissemination Standards Bulletin Board (DSBB) in October 2002. Appendix Table 4 provides an overview of Botswana’s dissemination practices compared to the GDDS.

2a. The Report on the Observance of Standards and Codes (ROSC)—Data Module contains the following main observations. Botswana has a macroeconomic statistical system that does not yet fully meet the needs of its policymakers and other data users. Botswana first undertook a ROSC data module assessment in October 2001, and it has used both the ROSC assessment and the GDDS detailed plans for improvement to guide the development of its capacity in the production of macroeconomic statistics. In the interval since the last assessment, there has been a strengthening of the relevance of the data (especially with regard to the responsiveness to users’ needs), the transparency of statistical policies and practices, the accessibility of data and metadata, and, on balance, some progress in the methodological framework that underpins the system. Staff of the statistical agencies exhibit a high level of professionalism and ethical standards. Overall, Botswana has a solid basis on which to prepare for the next stage in its statistical development, namely subscription to the more exacting requirements of the Special Data Dissemination Standard (SDDS).

2b. Indeed, there could be substantial benefits to Botswana from an early subscription to the SDDS, particularly as the economy continues to grow and evolve at a pace that risks outrunning the capacity of the statistical system to meet the needs of policymakers and other data users. However, Botswana would need to make a marked improvement in the periodicity and timeliness of its data in many areas in order to meet the standard. As well, it would need to begin compiling and disseminating certain statistics that are requirements for the SDDS, for example, statistics on general government, the International Reserves and Foreign Currency Liquidity Template (the Reserves Template), and key statistics on production and producer prices. Meeting these challenges would require high-level policy decisions to increase staff resources in the main statistics-producing agencies in a concerted and coordinated way, and to accept the high level of transparency regarding Botswana’s international reserves position that the Reserves Template requires.

2c. Improving the periodicity and timeliness of statistics should be an important goal, regardless of whether or not Botswana decides to subscribe to the SDDS; this will require more willingness to publish preliminary data—backed up by coherent and transparent revision policies and practices—than is the case at present. Other important goals should include strengthening source data and enhancing the domestic availability of metadata.

2d. Section II provides a summary assessment by agency and dataset based on a four-part scale. This is followed by staff recommendations in Section III. Practices compared to the GDDS are summarized in the Appendix. The authorities’ response to this report and a volume of detailed assessments are presented in separate documents.

3. In applying the IMF’s Data Quality Assessment Framework (DQAF), July 2003, the remainder of this section presents the mission’s main conclusions. The presentation is done at the level of the DQAF’s quality dimensions, by agency for the first two dimensions and across datasets for the remaining four.

4. With regard to prerequisites of quality, various laws and decrees authorize the Central Statistics Office (CSO), the Ministry of Finance and Development Planning (MFDP), and the Bank of Botswana (BoB) to collect, compile, and disseminate the relevant statistics. However, the legal framework does not provide an explicit mandate to BoB for compiling balance of payments statistics, or to the MFDP for disseminating fiscal statistics to the general public. In practice, responsibilities do not overlap, but cooperation among the agencies through informal agreement could be improved. While facilities and computing resources are broadly commensurate with the statistical program, the CSO and MFDP are seriously understaffed. This understaffing has adverse consequences for their ability to adequately perform the tasks required to produce the needed full range of national accounts, prices, and government finance statistics (GFS) in a timely way, let alone extend and further develop their statistical systems. All agencies have increased their efforts to monitor the overall quality of the statistical program and ensure that statistics remain relevant to users’ needs. They could, however, give more attention to the quality versus timeliness trade-off in their design and review of work programs, and place relatively more emphasis on timeliness. As to assurances of integrity, a combination of legislation and tradition supports the compilation and collection of statistics on an impartial basis. In all agencies, a culture of professionalism and objectivity is promoted and supported. The terms and conditions under which macroeconomic statistics are produced and disseminated are publicly available. However, there can be internal governmental access to all macroeconomic data except consumer and wholesale prices, prior to their public release, which is not always explicitly identified in domestic information sources. Staff of all of the statistics-producing agencies are held to a high ethical standard in the conduct of their work.

5. Concepts and definitions, with the notable exception of those underpinning the wholesale price index (WPI) and local government statistics, are in general methodologically sound, and broadly conform to internationally accepted standards. Monetary statistics follow the Monetary and Financial Statistics Manual (MFSM). Balance of payments statistics conform to the Balance of Payments Manual, fifth edition BPM5. Some changes are still needed to bring the national accounts fully into line with the System of National Accounts 1993 (1993 SNA). The GFS other than for local government generally conform to A Manual of Government Finance Statistics, 1986 (GFSM1986), while the consumer price index (CPI) conforms to the Consumer Price Index Manual 2004 (CPI Manual). Methodological problems, as well as issues of inadequate scope and faulty sectorization, seriously undermine the usefulness of the WPI. The institutional scope of the GFS does not extend to extrabudgetary institutions, consolidated central government, or consolidated general government. The scope of the remaining datasets is broadly in line with international standards with some relatively minor deviations. Sectorization and classification of data are generally sound, although there are problems in the national accounts data, which still mostly follow the classifications and sector breakdowns in the System of National Accounts 1968 (1968 SNA), and in GFS. The basis for recording generally follows the latest international guidelines.

6. The accuracy and reliability of macroeconomic statistics could be strengthened through reliance on better and more timely source data. For instance, data sourcing for balance of payments statistics could depend less on the International Transactions Reporting System (ITRS) with more recourse to alternative (principally survey) sources; the frequency of the main enterprise survey that underpins national accounts data could be increased; more detail is needed in the source data for monetary statistics, and the timeliness of BoB data should be increased sharply; the data used for compiling annual statistics for local government should be considerably more timely; and coverage for the WPI could be expanded. The techniques used to assess source data typically are sound. The agencies generally employ suitable statistical techniques in their operations, although the BoB could usefully strengthen its methods used for estimating missing data, adjusting import values, and estimating flows from stock data in the balance of payments statistics. Sound methods are generally used for assessing and validating intermediate data. Generally speaking, more— and more systematic—internal use could be made of revision studies to inform and strengthen statistical processes.

7. As to serviceability, datasets generally meet GDDS recommendations for periodicity, but fall short of the GDDS recommendations for timeliness in the case of government fiscal statistics and monetary statistics (the exception is the WPI, which meets neither of the recommendations). Datasets are mostly consistent within themselves and consistent over time; however, some inconsistencies are not reconciled among the datasets, for example, between balance of payments and national accounts statistics, and between government finance and monetary statistics. Revision policies and practices vary from dataset to dataset, but all would benefit from a more formalized approach, including conducting and publishing revision studies. Revised data are clearly identified.

8. The statistical agencies utilize their websites and publications to provide access to data and metadata. Data generally are clearly presented with a combination of charts and tables, and with a sufficient level of detail and historical information. More attention could be given to the use of advance release calendars, which should include projected release dates. Moreover, additional measures could be taken to facilitate access to metadata. In the case of government finance, monetary, and balance of payments statistics, metadata are available only from the IMF’s DSBB website. While the CSO does provide some brief methodological information in its publications and does make users aware of the DSBB source, it could also provide a link to it. The MFDP and BoB provide links to the DSBB on their websites, although the links are not particularly visible and are not associated directly with the data disseminated. The statistical agencies generally provide useful information—either on their website and/or in their publications—on contact persons for special assistance and queries.

II. Assessment by Agency and Dataset

9. Assessment of the quality of six macroeconomic datasets—national accounts, consumer price index, wholesale price index, government finance, monetary, and balance of payments statistics—was conducted using the DQAF, July 2003. In this section, the results are presented at the level of the DQAF elements and using a four-point rating scale (Table 1). Assessments of the prerequisites of data quality and the assurance of integrity (Dimensions “0” and “1” of the DQAF) are presented in Table 2aTable 2c. For each dataset, the assessment of methodological soundness, accuracy and reliability, serviceability, accessibility (Dimensions “2” to “5” of the DQAF) are shown in Tables Table 3af.

Table 1

Botswana: Data Quality Assessment Framework, July 2003—Summary Results

article image
Practice observed: current practices generally in observance meet or achieve the objectives of DQAF internationally accepted statistical practices without any significant deficiencies. Practice largely observed: some departures, but these are not seen as sufficient to raise doubts about the authorities’ ability to observe the DQAF practices. Practice largely not observed: significant departures and the authorities will need to take significant action to achieve observance. Practice not observed: most DQAF practices are not met. Not applicable: used only exceptionally when statistical practices do not apply to a country’s circumstances.
Table 2a

Botswana: Assessment of Data Quality—Dimensions 0 and 1—Central Statistics Office

article image
Table 2b

Botswana: Assessment of Data Quality—Dimensions 0 and 1—Ministry of Finance and Development Planning

article image
Table 2c

Botswana: Assessment of Data Quality—Dimensions 0 and 1—Bank of Botswana

article image
Table 3a

Botswana: Assessment of Data Quality—Dimensions 2 to 5—National Accounts

article image
Table 3b

Botswana: Assessment of Data Quality—Dimensions 2 to 5—Consumer Price Index

article image
Table 3c

Botswana: Assessment of Data Quality-—Dimensions 2 to 5—Wholesale Price Index

article image
Table 3d

Botswana: Assessment of Data Quality—Dimensions 2 to 5—Government Finance Statistics

article image
Table 3e

Botswana: Assessment of Data Quality—Dimensions 2 to 5—Monetary Statistics

article image
Table 3f

Botswana: Assessment of Data Quality—Dimensions 2 to 5—Balance of Payments Statistics

article image

III. Staff’s Recommendations

10. Based on the review of Botswana’s statistical practices, discussions with the data-producing agencies, and responses from data users, the mission has a set of recommendations. They are designed to increase further Botswana’s adherence to internationally accepted statistical practices and would, in the mission’s view, enhance the analytical usefulness of Botswana’s statistics. Some additional technical suggestions are included in the Detailed Assessments volume.

Cross-cutting recommendations

  • Strengthen cooperation between the MFDP, CSO, BoB, and other agencies to ensure timely compilation and exchange of data. The BoB should enjoy access to customs data, and the CSO should document what processing is carried out on trade data.

  • Monitor the consistency of the main macroeconomic datasets and reconcile differences regularly.

  • Give more emphasis to timeliness by using and disseminating more preliminary data.

  • Establish a comprehensive list of institutional units consistent with the 1993 SNA sectorization to be applied consistently across all datasets.

  • Establish and publicize direct links between the agencies’ websites and the DSBB to facilitate access to metadata. Keep metadata up-to-date

  • Support greater use of preliminary data by formalizing revision policies and implementing regular revision cycles.

  • Maintain a current Business Register.

National accounts

  • Increase the number of staff working on national accounts.

  • Document the methodology used to derive the 1994/95 benchmark estimates.

  • Undertake comprehensive surveys of enterprises at least every few years.

  • Introduce estimates for informal activities.

  • Bring classification and sectorization into line with the 1993 SNA.

  • Update the base year for the volume figures for GDP at least every five years.

Consumer price index

  • Include owner-occupied housing in the CPI.

  • Improve the overall quality adjustment techniques.

Wholesale price index

  • Increase the number of staff in the Price Statistics Unit, particularly those working on the WPI.

  • Develop concepts and definitions for Botswana’s WPI to meet the needs of data users. Basically, the CSO needs to determine if it wants to produce an output index (PPI), an Intermediate Consumption Index, or a Supply Price Index.

  • Develop a representative sample, including industries such as mining (gold and diamond industries), cattle production, etc.

  • Classify the WPI using internationally accepted standards such as the International Standard Industrial Classification of All Economic Activities (ISIC) and/or the Central Product Classification (CPC).

  • Conduct ad hoc surveys. Include own-account production of market goods for own final consumption in the WPI.

Government finance statistics

  • Compile and disseminate GFS for extrabudgetary institutions, consolidated central government, and consolidated general government, at least on an annual basis.

  • Formally assign responsibility for disseminating fiscal statistics to the public to the minister of the MFDP.

  • Allocate more human resources to the compilation and dissemination of GFS, and emphasize training in international methodologies. Reduce staff turnover.

  • Bring the classifications used for subannual and annual fiscal statistics on the operations of budgetary central government and local governments into conformity with the guidelines of the GFSM 1986. Consider adopting a migration path to the GFSM 2001.

  • Post the monthly summary cash-flow table on the MFDP’s website immediately after it becomes available. Compile a quarterly central government debt table and post it on the MFDP’s website. In this manner, Botswana would easily meet the GDDS timeliness standard.

Monetary statistics

  • Report the BoB’s preliminary monthly accounting data with a level of detail compatible with the recommendations of the MFSM to the Research Department, in electronic form and no later than 15 working days after the reference month.

  • Review the report forms of the ODCs to ensure the most complete and timely information for supervisory and statistical purposes. Establish electronic reporting systems to compile data from the ODCs.

Balance of payments statistics

  • Improve estimation methods and statistical techniques. Review benchmarks regularly, particularly for valuation adjustments to imports. Develop estimates for unrecorded trade and other missing data. Identify, as a matter of urgency, alternative sources for travel data.

  • Initiate planning for redeploying staff and other resources from processing bank reports on foreign exchange transactions to address other needs (such as surveys) and produce data of higher quality. Ultimately, terminate the banks’ foreign exchange reporting requirement.

  • Establish a formal intragovernmental agreement or change of the Bank of Botswana Act (BoB Act) to clarify the BoB’s responsibility for compiling balance of payments data.

Appendix

Table 4

Botswana: Overview of Current Practices Regarding Coverage, Periodicity, and Timeliness of Macroeconomic Data Compared to the General Data Dissemination System

article image