Information about Sub-Saharan Africa África subsahariana
Journal Issue
Share
Article

A Framework to Assess the Effectiveness of IMF Technical Assistance in National Accounts1

Author(s):
Gonzalo Pastor
Published Date:
December 2009
Share
  • ShareShare
Information about Sub-Saharan Africa África subsahariana
Show Summary Details

I. Introduction

This paper presents a framework for analyzing the effectiveness of technical assistance provided by the West Africa Regional Technical Assistance Center (AFRITAC West or AFW) in the area of national accounts using the Technical Assistance Information Management System (TAIMS), which is the Fund’s archive of core technical assistance information and documentation (see Box 1). AFW beneficiary countries are Benin, Burkina Faso, Côte d’Ivoire, Guinea, Guinea-Bissau, Mali, Mauritania, Niger, Senegal, and Togo. National accounts technical assistance projects are in place for all AFW countries, except Mauritania, thus providing a suitable country group for analysis. Also, for these countries, the national accounts projects have benefited from a systematic use of TAIMS; a practice that it is still lagging in most IMF regional technical assistance centers around the world.

The goal is to respond to the request by the AFW Steering Committee to report on technical assistance “ultimate outcomes” rather than on “inputs” (i.e., the number of national accounts missions fielded by the regional technical assistance center), which has been the practice to date. We proceed in two steps towards that goal. First, we use the Data Reports on the Observance of Standards and Codes (Data ROSCs) prepared for 13 African countries, including 3 AFW countries, as well as input from IMF country teams2 to AFW countries, to assess the data dimensions in need for improvement and thus, the “ultimate outcome” of the national accounts’ technical assistance project. Second, we reorganize the information contained in TAIMS using the Fund’s Data Quality Assessment Framework (DQAF) to make it comparable/consistent with the Data ROSCs’ data quality assessments. In principle, a strong consistency between the actual provision of technical assistance (as documented in TAIMS) and the reported data weaknesses (as reported by the Data ROSCs and IMF country teams) should increase the likelihood that the project’s “ultimate outcomes” would be attained within a reasonable time frame and budget.

Our analysis of TAIMS data concludes that the “ultimate outcome” of producing and disseminating robust national accounts is work in progress, with AFW’s technical assistance efforts mainly focusing on source data assessments and methodological issues underpinning the compilation of national accounts. The pending challenge, however, is to broaden the missions’ reach/scope and further support a more timely production and dissemination of national accounts data, which are pending challenges according to information contained in the Data ROSCs and input collected from IMF mission teams to AFW member countries. Striking the balance between data production and dissemination would be a challenge for data producers and users in AFW member countries. No doubt, without appropriate data sources, there is no technical assistance on methodology that could be successful. Yet, the international experience suggests that releasing data on a timely basis, with due reference to the data limitations, can raise the credibility of the national statistical offices and trigger a virtuous cycle, in which more data eventually results in better data, as data producers respond to pressures from data users.3

The Project Framework Summaries (PFSs) stored in TAIMS also suggest that the implementation status of measurable targets included in the national accounts technical assistance projects has varied across countries. Implementation has been strong in AFW countries which either have a relatively developed statistical system that gives countries a good starting point (Senegal), are rapidly catching up in terms of data production and dissemination after years of political crisis (Côte d’Ivoire), or are prompt to complete the updating of their national accounts supported by an array of new data sources (Niger). For other AFW country cases, the implementation status of major project targets has been weak and may remain a major technical challenge for years to come.

Moreover, the analysis points to a diverse technical assistance implementation record, not closely linked to the intensity of mission allocation. Several countries have managed to advance towards achieving their ultimate outcomes with relatively little support from AFW. This strengthens the case for enhanced TAIMS reporting on the circumstances surrounding the provision of AFW technical assistance, such as, for example, the role of other technical assistance providers and donors (e.g., multilateral/bilateral support from AFRISTAT, the African Development Bank, the World Bank, and other development partners).

This paper contributes to the debate on management by results regarding the provision of technical assistance by the Africa Regional Technical Assistance Centers (AFRITACs). It focuses on national accounts statistics, an area for which there exists a simple, yet rigorous, framework for data quality analysis: the Fund’s DQAF. However, for most other areas of IMF technical assistance (e.g., fiscal and financial markets’ development), standardized definitions of objectives/outcomes/indicators are still missing. Zeroing on a definitive approach to assess the effectiveness of the AFRITACs’ technical assistance delivery is also likely to require input and feedback from main stakeholders involved.

The structure of this paper is as follows. Section II analyzes the data users’ survey results and the IMF expert ratings on national accounts statistics, as reported in Data ROSCs prepared for 13 Sub-Saharan African countries, including 3 AFW member countries. It also appraises the IMF country teams’ replies to a questionnaire on national accounts issues. The analysis confirms that data users, IMF compilation experts, and IMF mission teams view the production and timely dissemination of national accounts data as the ultimate objective of technical assistance. Section III sketches the proposed framework to monitor progress towards achieving that objective using TAIMS. The framework is applied using AFW’s mission documentation stored in TAIMS for FY06-09. Section IV summarizes our findings and elaborates briefly on possible next steps.

Box 1:What is TAIMS?

The Technical Assistance Information Management System (TAIMS) is the IMF archive of core technical assistance information and documentation that is organized within a technical assistance project concept. The TAIMS’ database includes project details that are entered directly into the system (e.g., project’s ID and funding source) and information that is drawn from main technical assistance documents.

The standard documents stored in TAIMS include: (i) the Project Framework Summary (PFS), which is a rolling work plan detailing the project’s objectives and expected outcomes; (ii) the technical assistance missions’ briefing papers (BRP) that specify the missions’ purpose and main tasks consistent with the project objective; (iii) the Back-to-Office (BTO) reports that provide information on the extent to which the mission was successful in completing the tasks identified in the BRP; and (iv) the Project Assessment Report (PAR) which is completed by the country authorities a year after the completion of the project (country authorities may also complete interim PARs). The PARs allow the country authorities to provide information on the extent to which the projects’ objectives and outcomes were achieved, the sustainability of the project outcomes, and comments/suggestion for future technical assistance.

To date, the use of TAIMS for projects undertaken by the IMF Statistics department (STA) is limited to recording mission-related documents (BFRs and BTOs), including mission final reports. The system is also used, but not on a systematic basis, to record and update the objectives and outcomes PFSs.

II. Defining the National Accounts Projects “Ultimate Outcome

Two main sources were consulted to assess the “ultimate outcome” of AFW’s technical assistance program in the area of national accounts: (i) the Data ROSC prepared for African countries in recent years; and (ii) the IMF mission teams’ replies to a questionnaire on national accounts’ issues. Canvassing views from main data users and stakeholders is critical for determining the scope of technical assistance, its implementation timeline and the intensity of the technical support to AFW member countries. As shown below, all partner groups agreed on the need to work on the production of national accounts data with a minimum standard of accuracy and reliability, while, at the same time, ensuring timely data dissemination.

The information derived from the Data ROSCs is rich in two ways. First, it includes the IMF statistical experts’ assessments on the various aspects/dimensions of data collection, processing and dissemination in the respective countries, which are measured against best international practices. In essence, the experts’ assessments highlight the data dimensions in need of improvement, and thus the scope of future technical assistance. Second, for a selected number of cases, the reports include information on surveys that were launched by the Data ROSC missions to collect information/views from principal local users (i.e., private sector businesses, embassies, public agencies, and national and regional institutions) on the official statistics.

A survey of IMF country teams updates the Data ROSCs’ user surveys and highlights the concerns of economists working closely with policymakers in the context of surveillance and/or Fund programs. Currently, all AFW member countries, except Mauritania, have active PRGF programs. Financial programming exercises underpinning those programs rely, to a significant degree, on the available national accounts data.

A. The Data ROSCs: Experts’ and Users’ Views

Data ROSCs are one of the IMF instruments to support countries’ efforts to compile macroeconomic statistics according to best international practices.4 To date, Data ROSCs have been undertaken for about 90 IMF member countries. Data ROSCs contain IMF statistical experts’ assessments on main macroeconomic statistics within six data quality dimensions (Box 2). The data quality dimensions review institutional issues—such as transparency and professionalism in the data compilation process—as well as the professional rigor of the existing data compilation methods and dissemination practices.5

The Data ROSC experts’ assessments on each data quality dimension are further summarized in the form of ratings depending on the degree of adherence to internationally-accepted best compilation practices. An “O” rating indicates that best practices are observed in the country; “LO” means that best practices are largely observed; “LNO” indicates that best practices are largely not observed; and “NO” that best practices are not observed.

A total of 13 African countries, including 3 AFW member countries, have undertaken Data ROSCs to date. The data quality ratings assigned to national accounts data in all African countries point to a number of data weaknesses, which are further heightened in the case of AFW member countries (Table 1). In particular, data accuracy and reliability (DQAF 3) shows the highest need for improvement, followed by shortcomings with methodological soundness (DQAF 2) and data serviceability (DQAF 4):

  • National accounts’ data accuracy and reliability (DQAF 3) is impaired by poor quality of the source data used in the national accounts’ compilation. Limited information about the informal sector and the use of often-obsolete household budget and consumption surveys undermine the source data quality. There are also problems with the soundness of statistical techniques applied to derive main aggregates (e.g., use of outdated input-output tables for obtaining value added estimates, lack of adequate price indices to deflate different macroeconomic aggregates) and ad-hoc practices regarding data revisions (i.e., timing and analysis of the factors explaining the revised data).
  • Methodological soundness (DQAF 2) is weakened by a limited application across countries of concepts and definitions recommended by 1993 System of National Accounts (1993 SNA), in part because of source data limitations.
  • Data serviceability (DQAF 4), particularly data timeliness and periodicity, is hampered by the existence of long lags between the publication of different versions of national accounts (i.e., provisional, revised, and final estimates). This data weakness is particularly severe in the case of AFW member countries, whose data quality ratings for periodicity and timeliness are poor relative to other African countries. Because of the long delays in producing and publishing data, AFW users tend to resort to national accounts estimates prepared by data producers other than the national statistical offices using more aggregate (and usually nontransparent) compilation and estimation techniques.

Table 1.Africa: Country ROSC Ratings–National Accounts(By DQAF Data Quality Dimensions)
AFRITAC West CountriesOther African Countries
DQAFBurkina

Faso
NigerSenegalBotswanaThe

Gambia
KenyaMauritiusMozambiqueNamibiaSouth

Africa
TanzaniaUgandaZambia
0. Prerequisites of quality
0.1 Legal and institutional environmentLOLOLOOOLOOLNOOOOLNO
0.2 ResourcesLNOLNOLNOLOLNOLNOLOLNOLOLOLOLOLNO
0.3 RelevanceLOLOLOOLNOLOOLOOOOLO
0.4 Other quality managementLOLOLNOOOLOOLO
1. Assurances of integrity
1.1 ProfessionalismLOLOOOLOOOLOOOOLO
1.2 TransparencyLOLOLOLOLOLOOLOLNOOLOLOLO
1.3 Ethical standardsLOLOOOLOOOLOOOOO
2. Methodological soundness
2.1 Concepts and definitionsLOLOLOLOLNOOOOOLOLOLO
2.2 ScopeLNOLOLOLOLNOLNOLOLOOLOLOLOLO
2.3 Classification/sectorizationLNOLNOLOLNOOLNOOOLOLOOLO
2.4 Basis for recordingLOLOLNOLOLOLOLOLOLOLOLOLOLO
3. Accuracy and reliability
3.1 Source dataLNOLNOLNOLOLNOLNOLOLNOLNOLOLNOLOLNO
3.2 Assessment of source dataLOLOLOOLOOLOLOOOOLO
3.3 Statistical techniquesLNOLNOLNOLNOLNOLNOLNOLOLOOLNOLOLNO
3.4 Assess. validation of interm. data/outputsLOLOLNOLOLOLOLOLOOLOOLNO
3.5 Revision studiesLNONOLNOLONOLNOOLOOOLOOLNO
4. Serviceability
4.1 Periodicity and timelinessLNOLNOLOOOOOOOOOOO
4.2 ConsistencyLOLOLOLOOOLOLOLOOOOO
4.3 Revision policy and practiceLNOLOLOLNOLOLOOLOOOLOLOLO
5. Accessibility
5.1 Data accessibilityLNOLOLOLOLNOLOLOLOLOOOLOLNO
5.2 Metadata accessibilityLOLNOLOLOLNOOOLNOLOOLOLOO
5.3 Assistance to usersLOLOOOLNOLOOLNOLOOLNOOLO
Source: ROSC reports; www.imf.org.
Source: ROSC reports; www.imf.org.

Box 2:Data Quality Dimensions Assessed in Data ROSCs

Data quality dimensions assessed under the Data ROSCs review institutional issues, as well as the rigor of the existing data compilation methods and dissemination practices. All areas of analysis are classified under the DQAF, using specific codes (0 to 5). Institutional issues cover the legal environment underpinning the compilation of macroeconomic statistics and the professional quality of the individuals involved in the compilation process. Mythological rigor covers aspects of data concepts and definitions, source data quality, data serviceability and access to the public. Specific coverage under the various codes is presented below:

DQAF

Code
Data Quality DimensionSpecific Elements under Review
0Prerequisites of data qualityLegal and institutional environment is supportive of statistics; resources are commensurate with the needs of statistical programs; quality is recognized as a cornerstone of statistical work.
1Assurances of integrityProfessionalism in statistical policies and practices is a guiding principle; statistical policies & practices are transparent; statistical processes are guided by ethical standards.
2Methodological soundnessData concepts and definitions, data scope/classification/sectorization are in accord with international statistical practices. Flows and stocks are valued and recorded according to intl. accepted standards.
3Accuracy and reliabilitySource data available provide an adequate basis to compile statistics and portray the reality of the economy (i.e., assessment of source data); statistical techniques conform to sound statistical procedures; assessment and validation of intermediate data and statistical outputs are in place.
4ServiceabilityStatistics cover relevant information on the subject field; data timeliness and periodicity follow internationally accepted dissemination standards; data are consistent over time/across major data systems; data revision policies and practices follow regular procedures.
5AccessibilityStatistics are presented in a clear and understandable manner; forms of dissemination are adequate; data and metadata are easily available by users.

Data users’ surveys included in selected Data ROSCs add information to the statistical experts’ assessments. 6These surveys indicate that, except for South Africa, African users were mainly concerned about data serviceability and to a lesser extent about source data and methodology issues (Table 2). Timeliness was considered too long for GDP, fiscal and balance of payments data. There was also widespread criticism about the lack of information on official publication calendars for these statistics. Users’ views about source data and methodology issues—which were central to the Data ROSC experts’ assessments—were generally constrained by the reported poor data and metadata dissemination practices in all countries, except South Africa. Users’ comments and suggestions for improving data scope and detail varied across African countries depending on their stage of development of their statistical frameworks.

Table 2.Data ROSCs: Statistics Users’ Survey–Summary
Data

ROSC

Date
TimelinessScopeDegree of

Detail
FrequencyAdvanced

Publication

Calendar
Data/Metadata Accessibility
rankingcomment
AFW Countries
Burkina FasoMar-04too longinadequateinsufficientinadequatelittle info. availablesatisfactoryshould broaden dissemination
NigerJun-06too longinadequateinsufficientinadequatelittle info. availableunsatisfactoryshould broaden dissemination
SenegalNov-02too longbroadly adequateinadequateinadequatelittle info. availablemethodological notes are not very descriptive
Other African countries
KenyaOct-05too longadequatecould be improvedadequatelittle info. availableshould broaden dissemination
MozambiqueAug-05too longinadequateinsufficientinadequatelittle info. availableunsatisfactorymethodological notes are not very descriptive
South AfricaOct-01adequateadequateadequateadequatesatisfactorysatisfactoryseasonal adjustments & data revisions could be better explained
Source: IMF Data ROSC Reports.
Source: IMF Data ROSC Reports.

In the users’ view, improved data timeliness and periodicity according to a preannounced dissemination calendar is critical to develop a richer exchange between users and data producers, and an eventual quality improvement in national accounts compilation. A virtuous cycle would develop: a timely supply of data would create a demand for it that would, in turn, create pressures for better and more data production.

B. Survey of IMF Country Teams

A survey of IMF country teams conveyed the economists’ views/concerns over national accounts data for economic surveillance and IMF lending operations in AFW member countries. The survey, which was conducted during the last week of May 2009, recorded a very high response rate (see Annex I for the questionnaire and the country desks’ replies). The survey provided additional information on what should be considered the “ultimate outcome” of AFW’s national accounts technical assistance project.

Overall, the country teams suggested a need to establish a practical balance between the production and dissemination of national accounts along best international practices. Country desks are aware of the data accuracy limitations stemming from outdated base years (mostly dating from the 1990s) and the lack of regular survey data, which they address in their regular work through consistency checks with other macro data and consultations with institutions other than the national statistical offices. On data dissemination, all mission teams resort to receiving the national accounts data directly from the authorities rather than accessing the data through internet or other published sources/venues, which, in most cases have long publication lags or are nonexistent in AFW countries.

Country teams provided a number of recommendations for AFW on how to strike a practical balance between data production and dissemination:

  • AFW should develop local capacity so as to allow the country to produce and disseminate national accounts that meet minimum goals of accuracy within a reasonable resource envelope. Best international practices are not reachable in most AFW countries because it would be too resource-intensive, according to the economists.
  • Minimum standards could be set for data production and dissemination. On data production, these standards may possibly include a number of national accounts methodological benchmarks, including the development of consistency checks on the level and/or growth rates of critical national accounts variables (such as, for example, the economy’s savings rate, household final consumption, and investment to GDP ratios). Reportedly, the lack of consistency checks currently increases the data verification workload for country teams and results, at times, in significant data revisions.7
  • On data dissemination, the call is for setting objectives for data release lags. For example, the aim could be to produce within year t+1 reasonably sound preliminary estimates for year t, sound revised estimates for year t-1, and final estimates for year t-2. Such a state of affairs is, unfortunately, far from being in place to date.

III. A Monitoring Framework Using TAIMS

This section builds on the information collected for the Data ROSCS and the users’ surveys, as reported in section II. It sketches a framework to monitor progress towards producing and disseminating national accounts along best international practices, which is considered the “ultimate outcome” of the technical assistance project under review. Inputs for the analysis are the projects’ documentation stored in TAIMS, which we reorganize according to the six data quality dimensions assessed in the Data ROSCs (see Box 2 above) for reporting on the effectiveness of the actual technical assistance delivery. We also use the mission information contained in STA’s Regional Allocation Plans (RAPs) for FY06-09 to supplement TAIMS in assessing the intensity of the technical assistance effort—as measured by the number missions and nights spent in the field—in the AFW region.

The analysis proceeds in three stages:

  • First, we check the AFW missions’ scope (tasks and reported outcomes) against the data quality shortfalls identified by the statistical experts in the context of the Data ROSCs and by national accounts’ data users, including the IMF mission teams. A strong consistency between the missions’ scope (as reported in TAIMS) and the data quality shortfalls identified in the Data ROSCs and data users’ surveys should warrant a greater likelihood of achieving the project’s ultimate objectives. The latter being the production and dissemination of national accounts along best international practices.
  • Second, we review the available PFSs to assess the challenges in meeting the current technical assistance implementation deadlines for the various data quality dimensions assessed in the Data ROSCs.
  • Third, we report on the intensity of AFW’s technical assistance effort and the countries’ success in meeting the project’s “ultimate outcomes.”

Checking missions’ scope against reported data weaknesses

Upgrading a country’s national accounts is a major technical endeavor. The undertaking involves (i) the collection/updating of source data and (ii) changes/improvements in the national accounts’ compilation methodology in line with the 1993 SNA. According to the international experience, the combined implementation of changing the national accounts’ base year and implementing the 1993 SNA demands intense work (usually undertaken over a period of 5-6 years) and significant budgets. 8 Many data source upgrades9 usually evolve in tandem, while the implementation of methodological changes requires a careful consultation among main data producers and specialists, including international advisory groups. The latter provide input on other countries’ experiences on best ways to transit from outdated compilation methodologies embodied in the 1968 System of National Accounts (1968 SNA) to those recommended under the 1993 SNA.10

Overall, it appears that AFW has responded effectively to the challenge of developing robust national accounts in its member countries. A dedicated resident statistical advisor on national accounts, together with short-term experts, provided technical support to national statistical offices in the region. Also, the AFRITAC independent external evaluation concluded that the overall relevance of AFW statistics technical assistance effort was broadly “good”, notwithstanding the remaining compilation weaknesses, staffing issues in key statistical agencies, and challenges with donor coordination.

The information stored in TAIMS help us go beyond that general assessment. Specifically, we reorganize the information contained in TAIMS using the Fund DQAF to make it comparable/consistent with the data quality dimensions assessed under the Data ROSCs. The overall picture that emerges is that national accounts missions focused mainly on source data assessments (data accuracy and reliability) and data concepts/definitions issues (methodological soundness), while data timeliness (i.e., serviceability) and dissemination (i.e., accessibility) issues generally lagged behind in the missions’ agendas. On average, about 70 percent of the missions’ tasks dealt with source data and methodological issues (Table 3 and Annex III). Mali is an extreme case in which all missions recorded in TAIMS for FY06-09 focused on those two data quality aspects alone. By contrast, missions to Burkina Faso and Senegal emphasized issues of data serviceability and accessibility, possibly reflecting the countries’ efforts to address the low ratings received in these data categories during their Data ROSCs. Senegal’s upgrading efforts may also relate to the country’s objective of subscribing to the IMF Special Data Dissemination Standard (SDDS).11

Table 3:AFRITAC West: Cumulative Number of Mission Tasks, FY06-FY09(Percentage distribution of total mission tasks during period)
Data aspect\countryBeninBurkina

Faso
Cote

d’Ivoire
GuineaGuinea

Bissau
NigerMaliSenegalTogoAll

AFW

Countries
0. Prerequisite of quality41.237.50.00.020.04.80.017.633.316.0
1. Assurance of integrity0.00.00.00.00.00.00.00.00.00.0
2. Methodological soundness5.90.062.531.60.028.616.735.30.021.0
3. Accuracy & reliability52.912.525.063.250.052.483.329.466.747.9
4. Serviceability0.012.50.00.05.09.50.00.00.03.4
5. Accessibility0.037.512.55.325.04.80.017.60.011.8
Total mission tasks100.0100.0100.0100.0100.0100.0100.0100.0100.0100.0
Source: TAIMS and AFR staff estimates.
Source: TAIMS and AFR staff estimates.

Several factors explain the emphasis on source data assessments and methodological issues as reported in TAIMS. First, without appropriate data sources, there is no technical assistance on methodology and data dissemination that can be successful. Second, analysis of source data and methodology has responded to the completion of household and business surveys by the country authorities, which has not been easy to predict at times, and has required several consecutive missions to address different aspects of the source data collection and compilation processes. Third, the emphasis on source data and methodological issues is consistent with AFW’s country pages in the IMF’s General Data Dissemination System (GDDS; see Table 4). The GDDS country pages—usually produced by the national statistical offices—emphasize improvement plans for data collection and compilation, albeit without a commitment to disseminate those data along a strict calendar.

Table 4.AFRITAC West: GDDS Country Pages–Short-term Plans for Statistical Improvement(Number of tasks per data quality dimension)
Country\Data dimensionPrerequisites

of

Quality
Assurance

of

Integrity
Methodological

Soundness
Accuracy

&

Reliability
ServiceabilityAccessibility
BeninXX X X XX X X XX X
Burkina FasoX XXX XXX
Cote d’IvoireXXX X X
GuineaX XX
Guinea BissauX XX X X X
MaliX XXX
NigerX XX X X X X XXX
SenegalX XX XX
TogoXX X X
Source: IMF GDDS country pages.
Source: IMF GDDS country pages.

Yet, the need to simultaneously address source data and methodological issues, as well as timely data dissemination is supported by the Data ROSC experts’ assessments, input from main data users, and the international experience. Main arguments in this regard are the following:

  • AFW member countries need to catch up with data serviceability practices in other African countries. In particular, data periodicity and timeliness elsewhere in Africa proceeds along best international practices (Table 1), while best practices in data serviceability in the AFW region are “largely not observed.” Notably, data periodicity and timeliness were main concerns for data users, including the Fund mission teams.
  • The international experience suggests that timely data dissemination narrows the scope for data manipulation. Published data, supported by a clear methodology that emphasizes the scope and limitations of the data, raises the credibility of the data producing agencies, reduces faulty data interpretations, and increases the usefulness of the data for economic analysis.12
  • Also, timely data release triggers data users’ demands that put pressure for more and better data. This is the virtuous cycle of data production and dissemination observed in countries around the world.
  • Finally, there is scope for country differentiation. The timeliness of the data publication should be consistent with the data producers’ resource envelope, although a data publication calendar should be in place and binding in all cases. Experience suggests a trade-off regarding publication lags: unduly short lags may lead to inaccurate/unreliable data and trigger numerous data revisions. Unduly long lags generate obsolete data that are not useful for decision-making.

Checking implementation deadlines

In this section we compare the expected implementation deadlines for technical assistance, as reported in the available PFSs, against the actual status of the various tasks. While the TAIMS information about the current standing of key milestones is limited, a number of preliminary conclusions based on those data and our earlier analysis are warranted:

  • According to the PFSs, deadlines for data serviceability and accessibility are mostly due in the second half 2009 (Table 5), a time bound that appears tight for all AFW member countries, except Côte d’Ivoire and Senegal. Côte d’ Ivoire is rapidly catching up in terms of national accounts data production and dissemination after years of domestic political crisis. Technical assistance to Senegal has mainly focused on the development of quarterly national accounts. According to the information stored in TAIMS, the Senegal authorities are broadly on target to deliver a timely publications of these accounts, which is consistent with their intention to subscribe to the Fund’s SDDS.
  • Publicly available information, and the evidence reported in TAIMS, highlights the “catching up” effort needed to meet national accounts’ data dissemination challenges in AFW member countries. According to our preliminary assessment (see Box 3 and Table 7), by end-2009, except for Senegal, all other 9 AFW countries would need to publish two or three years of final or semi-final national accounts data (2006 to 2007) and one year of preliminary national accounts (2008). This catching up effort would be consistent with the “rule of thumb” proposed by IMF country teams of producing within year t+1 reasonably sound preliminary estimates for year t, sound semi-final estimates for year t-1, and final estimates for year t-2.
  • The PFSs indicate that very few measurable targets within the grand national accounts’ technical assistance project have been completed to date (Annex IV). These successful cases generally apply to AFW countries that have relatively advanced statistical systems (Senegal and Côte d’ Ivoire) or are very close to revise their national accounts’ base year (Niger). Examples of these accomplishments include staff training in the area of quarterly national accounts (Senegal), completion of data survey questionnaires (Niger), availability of robust annual (Niger) and higher frequency (Senegal) source data for national accounts compilation, and completion of household expenditure surveys to support a direct estimate of household final consumption in the national accounts (Côte d’ Ivoire). The remaining measurable targets are work in progress, mainly affecting those countries with significantly outdated national accounts base years. Pending common challenges across countries include the allocation of sufficient human and financial resources to produce national accounts data (Benin, Guinea, Guinea-Bissau),13 staff training on a software accounting platform developed by the European Commission and the government of France (Benin, Côte d’Ivoire, Guinea-Bissau, and Mali), and the strengthening of compilation and data validation techniques–including adjustments for missing observations and undercoverage of production activities–for international trade data and the government sector statistics (Benin, Guinea-Bissau, Niger, and Mali). A measurable target that is still pending in Niger, Guinea-Bissau and Mali refers to the development of data surveys on their mineral and energy sectors.

Table 5.AFW: Technical Assistance Implementation Deadlines By Data Quality Dimensions a/(As reported in Project Framework Summaries)
Prerequisite

of Quality
Methodological

soundness
Accuracy &

Reliability
ServiceabilityAccessibility
BeninMid-09Dec. 09Q4-09Q4-09
Côte

d’Ivoire
Mid-09Q2 & Q4-09Q3-09June 09
GuineaMid- 2010June 2010June 2010
Guinea-

Bissau
April 2010Dec. 2010Q2 & Q4-09Q3/Q4-09
NigerDec. 2010II half 09;

end-2010
Q4-09Oct. 09
MaliDec. 08Dec. 08June 09
SenegalDec. 09May 09Q2 & Q3-09Dec. 09
Source: Annex IV.

No PFSs are available in TAIMS for Burkina Faso and Togo.

Source: Annex IV.

No PFSs are available in TAIMS for Burkina Faso and Togo.

Table 6:EIU–Country Reports: National Accounts’ Data’s Timeliness & Information Source
Date of EIU ReportLatest “actual”

GDP data (year)
“Estimated GDP data”

(years, source)
Other referred sources on national accounts data
BeninApril 2009(2005)(2006-08, EIU)Association interprofessionelle du coton; IMF
Burkina FasoMay 2009(2006)(2007-08, EIU)IMF
Côte d’IvoireMay 2009(2007)(2008-10, EIU)WAEMU; Direction Générale des Douanes; Direction de la Conjoncture et de la Prévision économique; IMF
GuineaJune 2009(2005)(2006-08, EIU)IMF
Guinea-BissauApril 2009(2006)(2007-08, EIU)IMF
MaliMay 2009(2006)(2007-08, EIU)BCEAO; Ministry of Agriculture; IMF
MauritaniaApril 2009(2006)(2007-08, EIU)IMF
NigerMay 2009(2006)(2007-08, EIU, IMF)IMF
SenegalMay 2009(2008)(2009-10, EIU)Agence Nationale de la Statistique et de la Démographie, Ministry of Economy and Finance; IMF
TogoApril 2009(2006)(2007-08, EIU)BCEAO; WAEMU; Afristat; IMF
BotswanaMay 2009(2008)(2008-10, EIU)Central Statistics Office; Bank of Botswana, Botswana Financial Statistics, Annual Report; IMF
The GambiaApril 2009(2007)(2008, EIU)IMF
KenyaMay 2009(2007)(2008-10, EIU)Central Bank of Kenya; IMF
MauritiusMay 2009(2008)(2009-10, EIU)Central Statistics Office; Ministry of Tourism, Leisure and External Communications; Ministry of Finance and Economic Development; IMF
MozambiqueMay 2009(2007)(2008-10, EIU)UN Food and Agriculture Organization; IMF
NamibiaMay 2009(2008)(2009-10, EIU)Bank of Namibia; Central Bureau of Statistics; Irwin, Jacobs, Greene/Institute for Public Policy Research, Windhoek; IMF
South AfricaMay 2009(2007)(2008-10, EIU)Statistics South Africa, South African Reserve Bank; IMF
TanzaniaMay 2009(2007)(2008-10, EIU)Barrick Gold; UN Food and Agriculture Organization; Bank of Tanzania; IMF
UgandaMay 2009(2007)(2008-10, EIU)Bank of Uganda; Uganda Bureau of Statistics; IMF
ZambiaJune 2009(2007)(2008-10, EIU)Ministry of Agriculture and Cooperatives, Bank of Zambia; IMF
Source: EIU Country reports, various.
Source: EIU Country reports, various.
Table 7.AFW: Number of National Accounts’ Missions & Mission Nights, FY06-FY09 1/
Number ofDistribution by
MissionsMission

Nights
MissionsMission

Nights
Benin3316.17.6
Burkina Faso4258.26.1
Cote d’Ivoire4308.27.3
Guinea88616.321.0
Guinea-Bissau76514.315.9
Mali65312.213.0
Mauritania 2/
Niger3276.16.6
Senegal106020.414.7
Togo4328.27.8
Total49409100.0100.0
Average5.445.4
Source: STA Regional Allocation Plans (FY06-FY09).

Includes missions undertaken by long-term and short-term experts.

Currently not eligible for IMF technical assistance.

Source: STA Regional Allocation Plans (FY06-FY09).

Includes missions undertaken by long-term and short-term experts.

Currently not eligible for IMF technical assistance.

Box 3:Tracking the National Accounts’ Data Timeliness & Information Source in EIU Country Reports

We reviewed the latest available country reports produced by the Economic Intelligence Unit (EIU) for a group of African countries, including all AFW member countries and those undertook Data ROSC examinations in recent years. The EIU country reports contain regular and up-to-date analysis of economic and financial developments for countries around the world. The EIU’s publication cycle is almost identical for all country reports, thus proving a helpful benchmark for cross country comparisons for data availability at a point in time.

The country reports’ review confirms the difficult situation with data timeliness in AFW member countries (see Table 6):

  • As of May/June 2009, except for Senegal and Côte d’Ivoire, “actual” annual national accounts data for AFW member countries were significantly outdated, dating from 2005 (2 countries) or 2006 (6 countries).
  • Data timeliness for the rest of African countries reviewed was much better, with “actual” GDP data dating from 2007 (7 countries) or 2008 (3 countries).
  • GDP “estimates” prepared by the EIU covered any data gaps (for the recent past or short-term future). No information on the EIU’s national accounts estimation methodology was included in the country reports.
  • For most AFW member countries, the IMF (i.e., IFS or the Fund’s Direction of Trade Statistics) was the only additional data source referred in the EIU country reports. This contrasts with the multiple domestic data sources referred in other African cases.

Checking for the intensity of TA delivery

Mission data compiled in TAIMS and the RAPs for FY06-09 confirm AFW’s vast effort to support countries’ initiatives to upgrade their national accounts (Table 7). During the last four fiscal years, there were 49 national accounts missions to the region, with a total of staff nights in the field reaching about 410. However, the mission distribution was not even across countries, with Benin, Burkina Faso, Côte d’Ivoire, Niger, and Togo receiving relatively less missions than the rest of AFW countries.

There seems to be no clear consistency between the intensity of the project (as measured by the number of missions and mission nights) and the reported effectiveness of the technical assistance delivery. Countries like Côte d’Ivoire and Niger received limited technical assistance support from AFW, but are reportedly prompt to attain a timely production and dissemination of national accounts (Côte de Ivoire) or update the national accounts’ base year (Niger). Senegal is a case in which an intense technical assistance effort (10 missions and 60 nights in the field) has delivered important results in the production of quarterly national accounts and the development of institutional sectors accounts. By contrast, the outcome has been much limited in countries like Guinea, Guinea-Bissau and Mali, notwithstanding AFW’s intense technical assistance effort. Mali has had the added advantage of being AFW’s host country, which, in principle, should have speeded up mission delivery processes and eased feedback between the country authorities and the local technical assistance experts.

While the PFSs or other information stored in TAIMS provide limited information about the circumstances surrounding the provision of technical assistance, two factors may explain the diverse results across countries:

  • Countries like Guinea, Guinea-Bissau, and Togo are currently considered fragile states in which civil war, political instability, and weak government institutions may have hampered technical assistance implementation. Côte d’ Ivoire is also a fragile state, although it has apparently managed to address data weakness following progress in resolving its civil conflict.
  • Donor support may have played a key role in funding activities that complemented the technical advice provided by AFW. A timely funding of surveys, censuses and other data sources may have eased technical assistance implementation. Further country analysis in this regard seems warranted to highlight the multilateral/bilateral support from AFRISTAT, the African Development Bank, the World Bank, and other friendly development partners.

Looking to the future, a regular reporting of the external environment surrounding the provision of technical assistance may provide valuable signals on the likely record of technical assistance implementation. PFSs may benefit from a timely reporting/updating on the work of other technical assistance providers. This information may, in turn, influence the sequencing of AFW’s technical assistance provision to a country.

IV. Concluding Remarks

This note represents an effort to analyze the effectiveness of technical assistance provided by AFW in the area of national accounts using TAIMS. The challenge has been to report on “ultimate outcomes” (i.e., the production and dissemination of national accounts statistics along best international practices) rather than on “inputs” (i.e., the number of national accounts missions fielded by the regional technical assistance center), as has been the case to date.

The analysis has benefited from the national accounts’ mission documentation stored in TAIMS, information included in the Data ROSCs launched for African countries in recent years, as well as input received from the IMF country teams to AFW countries. The information collected from the Data ROSCs and the country teams helped us define the project’s “ultimate outcome,” which we then used as a benchmark to gauge the effectiveness of the actual technical assistance delivery.

Central to the analysis has been the application of the Fund’s DQAF to organize the information collected in six specific data quality dimensions. These data dimensions range from institutional factors surrounding the production of national accounts data to quality factors referring to data compilation/processing/dissemination processes in place in the various AFW countries. As such, the DQAF provided a flexible structure for our qualitative assessment of AFW’s technical assistance work on national accounts.

Our analysis of TAIMS data suggests that the “ultimate outcome” of producing and disseminating robust national accounts is work in progress, with AFW’s technical assistance efforts mainly focusing on source data assessments and methodological issues underpinning the compilation of national accounts. The pending challenge, however, is to further support a more timely production and dissemination of national accounts data, which are main data weaknesses according to information contained in the Data ROSCs and input from IMF mission teams to AFW member countries.

While the national accounts missions’ emphasis on source data assessments and methodological issues is indeed critical, the international experience recommends that the production and timely dissemination of national accounts should advance in parallel. This joint approach to statistical development would also help AFW member countries catch up with other African countries in terms of data periodicity and timeliness. Currently, all IMF country teams to AFW countries rely on the country authorities to collect latest national accounts data for purposes of surveillance and IMF lending operations. However, a timely availability of the same data to other data users is not guaranteed, with external analysts (e.g., The EIU country reports) resorting to their own national accounts calculations in the absence of official data.

Striking the balance between a timely production and dissemination of national accounts data would be a challenge for data producers in AFW member countries, albeit with a number of benefits. For one, published data, supported by a clear methodology that emphasizes the scope and limitations of the data, raises the credibility of the data producing agencies, reduces faulty data interpretations, and increases the usefulness of the data for economic analysis. A timely data release usually triggers data users’ demands that put pressure for more and better data. This is the virtuous cycle of data production and dissemination observed in countries around the world.

The analysis of TAIMS also suggests that the implementation status of measurable targets included in the national accounts technical assistance project appears to be stronger in AFW countries which either have relatively developed statistical system that improves the countries’ initial conditions (Senegal), are rapidly catching up in terms of data production and dissemination after years of political crisis (Côte d’Ivoire), or are prompt to complete an updating of their national accounts using newly-available data surveys and censuses (Niger). For other AFW country cases, the implementation status of major project targets may remain weak for years to come.

Finally, the analysis points at the diverse technical assistance implementation record, notwithstanding the intensity of the mission allocation. Several countries have managed to advance towards achieving their ultimate outcomes with relatively little support from AFW. This strengthens the case for enhanced TAIMS reporting on the circumstances surrounding the provision of technical assistance, such as, for example, the role of other technical assistance providers and donors in areas supporting the implementation of AFW technical assistance.

Looking to the future, a number of observations emerge from the analysis:

  • Conducting a thorough analysis of technical assistance in all AFRITACs would require a solid effort from the centers to input the missions’ documentation in TAIMS. Currently, the information in TAIMS is rather limited or absent for a number of countries.
  • AFW mission teams could do a better job in defining their tasks in the missions’ briefing papers. Mission tasks are currently recorded in rather general terms, thus hampering a clear-cut classification according to the DQAF proposed in this note. A possible strategy could be for the AFRITACs to develop/define a number of “menu tasks” based on their country/regional experiences. Standardizing the mission tasks would ease the tracking over time of the missions’ tasks and ultimate objectives, albeit with an initial time investment by the AFRITACs.
  • PFSs need to be developed and updated on a regular basis for all AFW countries to convey information to monitor the evolution and assess the effectiveness of the national accounts projects in place. PFSs should include regular updates on other technical assistance providers and donors operating in the same topical area.
  • Consolidating the TAIMS data into the proposed analytical format presented above may require cooperation with information technology experts to automatically extract the missions’ tasks information listed according to the DQAF. This would reduce the time spent in manual manipulations of the mission briefs’ information.
  • Data ROSC provide important benchmarks on the various quality aspects of data collection, processing, and dissemination. To date, only a limited number of African countries have undergone these comprehensive data assessments. Increasing the number of Data ROSC for Africa would provide observes with valuable information to assess the effectiveness of technical assistance provided by the AFRITACs.

This paper has offered input to the debate on management by results regarding the provision of technical assistance in national accounts by the AFRITACs. The proposed framework is an effort to structure the debate within the rather simple, yet rigorous, context of the Fund’s DQAF. Zeroing on a definitive approach to assess the effectiveness of the AFRITACs’ technical assistance delivery in areas other than macroeconomic statistics (e.g., fiscal and financial sector management) would require the development of standardized definitions of objectives-outcomes-indicators, which are still missing. Such an approach will also require input and feedback from main stakeholders involved in the delivery and implementation of the technical advice.

Selected Bibliography

    International Monetary Fund (various dates & countries), Reports on the Observance of Standards and Codes, http://www.imf.org/external/np/rosc/rosc.asp?sort=date.

    • Search Google Scholar
    • Export Citation

    Olinto RamosRoberto, GonzaloPastor and LisbethRivas(2008), “Latin America: Highlights from the Implementation of the System of National Accounts 1993 (1993 SNA),” IMF Working Paper, WP/08/239, October, 51 pages.

    • Search Google Scholar
    • Export Citation
Annex I: IMF Country Teams’ National Accounts Questionnaire and Replies
Question\reply

(National

accounts base

year)
Benin

(1985)
Burkina Faso

(1999)
Côte d’Ivoire

(1996)
Guinea

(2003)
Guinea-

Bissau

(1986)
Mali

(1987)
Niger

(1987)
Senegal

(1999)
1. Would you agree that the ultimate goal of the project should be the production and timely dissemination of national accounts along best international practices?YesYesYes. The goal is reachable for CIV. Current problems in NA data production or dissemination derive more from the years of domestic political crisis than from a lack of capacity.The ultimate goal is too ambitious.YesYesYesYes, in general. In the specific case of Senegal, the goal is to start disseminating reliable quarterly national accounts (QNA) by the end of the first quarter of 2010.
2. If not, which should be the TA’s ultimate goal in the area of national accounts? How is AFW servicing the country authorities reach that goal?N.A.N.A.N.A.The ultimate goal should be to help build domestic capacity so as to allow the country to timely and periodically produce and disseminate national accounts that meet minimum standards of accuracy within reasonable envelope resources. Best international practices are not reachable by most AFRITAC countries, because it would be too resource- intensive. For instance, quarterly national accounts are out of reach in many AFRITAC countries.N.A.N.A.N.A.N.A.
3. Would you give the same importance to (i) the production and (ii) the timely dissemination of sound national accounts?Yes, the same importance for production and the timely dissemination.No; Production is much more important than disseminationMore importance should be given to timely dissemination, even though progress has to be made on both aspects. CIV is currently catching up, after years of delays. Two years of definitive national accounts (2004 and 2005) were finalized in 2008. This catching-up effort in production and dissemination should be supported, with a focus on dissemination.The aim should be to produce within year t+1 reasonably sound preliminary estimates for year t, sound semi-final estimates for year t-1, and final estimates for year t-2.YesYes, both essential.Production and disseminatio n are all important. Good production is, however, the basis for disseminatio n. In a resource-constraint situation, I will tend to give more weight to the production side than to the disseminatio n side.These two objectives should be pursued in parallel.
4. What, in your view, are the main shortfalls of the national accounts as currently produced? How do you address those shortfalls in your country work?Inadequate resources and weaknesses in methodology hamper the accuracy and reliability of the national accounts. Most data are available only on a yearly basis and are not produced for higher frequencies (Semi-annual or Quarterly). These weaknesses have a significant impact the mission’s work, by increasing the workload on mission and in HQ and by leading to delays in the production of reports. Furthermore the data sent by the authorities go through many stages of revisions before they are confirmed by the authorities. Sometimes these revisions are very significant raising serious questions about the methodology.National account data is not reliable. The office in charge of the production of national accounts (DGEP) does not have data collection capabilities.Some of the main shortfalls have already been flagged to AFW: problems with GDP deflators and with CPI components (weights should be updated and share of imported goods should also be taken into account). Other shortfalls are the absence of primary statistical surveys taking into account the former rebel zone.The main shortfalls are the absence of good primary statistical surveys on which national accounts should be built. This is particularly true of Balance of Payment statistics and industrial and services production surveys. This may explain that despite the recent revised base year and improved methodology, revised estimates for national accounts show unrealistic developments, notably in the areas that are often balancing items such as savings.National accounts data is currently not based on survey data. That is why we use 2002 as a base year and estimate annual subsectori al real growth rates (and annual subsectori al deflator growth rates) to derive annual real (and nominal) GDP data. The authorities are using the same methodolo gy and therefore have the same data.(i) Long production lags; (ii) internal inconsistencies in the data; (iii) no expenditure- side accounts; (iv) key “satellite accounts” not readily available (e.g., for gold sector); (v) major inconsistencies in the transition from factor cost to market prices (i.e. indirect tax estimates); (vi) calculation of deflators is problematic; (vii) reconciliation between the supply and demand side is through inventories; (viii) BOP data are incomplete (external assistance and financing are not comprehensive, while coverage of imports is more complete and timely leading to large errors and omissions).The main shortfall of the national accounts as currently produced is the base year which is 1987. Advising the National Statistics Office (NSO) to adopt another base year. With the assistance of AFRITAC West, they are working on 2006 as the new base year.The annual periodicity of the national accounts appears to be the main shortfall, along with the limited estimate of the scope of unrecorded activities. These shortfalls do not hinder our country work.
5. Which is the national accounts’ base year in your country? If the base year is rather old, is that an issue for you? Why? How do you address this problem?The base year is 1985. It is indeed old and a more recent base year (such as 2000) would be preferable.1999. So far it has not been an issueThe current base year in N. A is 1996. This is obviously too old. A new base year (2007 or 2008) should be adopted soonThe base year was recently changed to 2003, and revised (final) national accounts were produced for 2003-2005. Given the large fluctuations in certain prices since then, one should consider preparing a new base (2008) and chain the different price bases to maintain long time series (retropolation before 2003 is not available).The base year in the authorities ‘ series is 1986. We, instead, use 2000 and therefore simply rebase the series1987. Among the many other problems, this one doesn’t stand out. Moreover, the authorities are working on an update to 2000.The base year in the national accounts of Niger is 1987 which is pretty old. We have rebased to 2000 the computation of some national accounts categories (GDP deflator, Real GDP…)The base year is 1999 (changed from 1987 following the 2002 ROSC). This has not been an issue for our country work so far.
6. Has the TA provided by AFW during the last 2-3 years led to improvements in the national accounts production? For example, have either the nominal GDP levels or the growth rates of the economy been revised as a result of the TA work? Are there other elements of GDP being affected by the TA provided (e.g., share of investment in GDP; size of net exports)? Other issues (e.g., improved coordination among source data producers)?Efforts to address these shortcomings are ongoing. Benin participates in WAEMU’s harmonization of statistical methodologie s and in the GDDS project for AFRITAC West countries to implement the 1993 SNA. A statistical register and an industrial production index are being developed, but implementation is not advancing as expected and the August 2008 West AFRITAC mission discussed with the authorities the necessary steps to accelerate the compilation of the revised accounts and disseminate them by end-2008.The production of national account data has improved in recent times. However, no revisions to GDP and its element have been made as a result of recommendation s of AFW TATA provided by AFW during the last 2-3 years has helped support the ongoing catching-up in NA production. However, there have been significant delays in implementing specific TA ecommendation nsTA has aimed at producing data using appropriate methodology and new base year. However, the outcome has been disappointing. Capacity building is doubtful given the high turnover and lack of qualified personnel and there is no evidence that Guinea is becoming self-reliant.There was a recent TA, but we have not seen the outcome yet. The authorities preferred to wait until the finalization of the data before passing them on to us. Thus, we are planning to integrate the new data during our next mission (Aug/Sept 2009).Turnover on the team has erased our institutional memory on this question.Technical assistance has yet not led to revisions of growth rate or major aggregates in the national accounts. However, implementati on of ERETES with the new base year will lead to data and growth rate revisions for some years (2005, 2006, 2007…). Exports and imports will also be revised in the national accounts data to make them consistent with those of the BOP data. TA has also helped make progress in implementin g the 1993 System Of National Accounts (1993 SNA).Over the past two years, TA provided by AFW has followed a solid timetable for developing institutional sector accounts and QNA. This ongoing work is promising.
7. What are the main shortfalls of the national accounts as currently disseminated? How do you address those shortfalls in your country work?They are not disseminated along the best international practices.National accounts are not disseminated in a timely manner. They are made available to staff and the public during missions.Provisional data are now disseminated on a reasonable timeline (1-2 years after the end-of-year). Definitive national accounts were produced and disseminated with increasing delays during the crisis. The situation is now improving.Latest (provisional) data made available to staff date from 2006. Questions on the quality of some series are yet to be answered. Past experience on Guinea and other fragile low-income countries indicate that deadlines for the production of national accounts are regularly breached. The case has to be made at the highest level (Minister of Economy) that timely available national accounts data are key to design sound macroeconomic policies.(i) Given that we are only using estimates, data reliability and accuracy are the main problems (see point 5 above). (ii) Besides, the pre-2002 data is fairly consistent -especially nominal GDP in high-inflation years. (iii) So far, the informal economy is not taken into account at all.(i) By the time final accounts are disseminated, they are so old we are not clear as to who is using them; (ii) the statistical office is mainly a collector of sectoral information and compiling institution–as such it lacks some independence and quality control of the data provided by sectoral ministries cannot be achieved.With the creation of the Website of the National Office of Statistics, data disseminatio n has improved. Most of the data disseminate d are still low frequency. Improvemen ts in the process of data collection and treatment by strengthenin g capacities would help to produce high frequency data and data dissemination.Regular exchanges with the authorities (in particular, the forecasting agency which provides updated estimates at the time of missions) allow us to cope with the lengthy dissemination (9 months) of national accounts.
8. Has the TA provided by AFW during the last 2-3 years led to improvements in the dissemination of national accounts? Are the data more timely and regularly distributed to users than before?No. No.Data is disseminated more timely than before. Cannot tell whether it is because of previous AFW TA.TA of AFW has accompanied the catching up of NA dissemination. However, further progress is crucialDon’t know.See point 6 above.See answer to #6. The presentation of the 2007 BOP data by the BCEAO was done last week (nearly a one year and a half lag).Authorities support that TA provided has helped disseminate national accounts data in a more timely and regular manner than before.See answer to question 6.
9. Are the latest national accounts data published in IFS or the national statistical office’s website? Do you ever look for those data in IFS or the internet, or do you get the data directly from the authorities?They are published in IFS with a significant delay of more than year. We get the data directly from the authorities.National accounts data are published in the IFS. The statistics bureau does not work on national account data. We get the data directly from the authorities.We get most of the data directly from authorities.No. Data are provided directly by the authorities.We directly get the data from the authorities. As a matter of fact, we even prepare the final data with them.Published in IFS (no local website), but data not timely: we get directly from authorities. There is, however, regular publication on a monthly basis of recent economic developments.No, the latest national accounts are not in the published section of IFS. GDP data have been published in IFS up to 2001. We get the data directly from the authorities. We consult sometimes the National Office of Statistics website.The latest (2007) data are published in IFS and on the national statistical office’s website. We may incidentally look at the national statistical office’s website, but we usually get the data directly from the authorities.
10. In your view, what could be done differently by AFW to ease your country work in terms of national accounts’ analysis? Have you transmitted those ideas to AFW in the context of reviewing the mission briefs and/or TA reports? If not, why not?The AFW could work with the authorities to improve the quality and reliability of the national accounts. Yes.More technical assistance in the production of national account data. Emphasis should be placed on the development of data collection/analysi s systems and capabilities.AFW could also help consolidate the links between the INS (Institut National de la Statistique-producer of NA data) and the DP (Direction de la Prevision – Economic forecast and analysis).AFW should not limit its work to help produce national accounts number. It should also help national account producers to produce an annual report on (i) analyzing each vintage year, explaining revisions and limitations.We do not see particular short-term wins; we are counting on AFW to improve the overall quality of data in the medium term. Coordination among other donors supporting the strengthening of the statistical system should be investigated.To ease our work on the national accounts, we believe that AFW should help develop data production with high frequency (quarterly GDP), leading indicators, indicators on employment and unemployment… We have transmitted some of these ideas, for instance making consistent the national accounts data on exports and imports with those of the BOP. We haven’t had an opportunity to comment on other issues.The work undertaken by AFW on national accounts is promising. We are looking forward to its outcome in 2010. We have occasionally made some suggestions in reviewing mission briefs. We suggest that AFW do not hesitate to routinely contact the Senegal desk to exchange information.
11. What could AFW do differently to better serve the country authorities’ interest regarding national accounts’ production and data dissemination?Help the authorities to accelerate the implementatio n of the aforemention ed projects (see 6).AFW could work directly with country authorities to implement TA recommendations.AFW missions in CIV should be longer (a couple of weeks instead of a few days), in order to facilitate identification of internal key issues or bottlenecks in the production and dissemination of data. Moreover, AFW should coordinate its TA with other donors (EU, bilaterals…), or even take the lead in TA coordination.AFW should also establish a few performance indicators and provide incentives (for instance participation in seminars and training inside and outside the country) for each significant improvement in these indicators.No particular suggestions. On a short term horizon, AFW could help the authorities in building the satellite accounts of the cotton and gold mining sectors with a view to better understanding the macroeconomi c impact of these sectors (on the BOP, real, budget aspects of these sectors).Discussions with the authorities indicate that strengthenin g capacity of staff working in the production of national accounts data would be welcome.AFW’s TA on QNA is what the authorities want as it squares with their objective to subscribe to the SDDS. Looking forward (beyond the 2010 dissemination date of the first QNA), it will be important for AFW to consult closely with the authorities on next steps.
Annex II: 1993 SNA Main Methodological Changes from 1968 SNA14
1993 SNA1968 SNA
Valuation of output and value added is done at basic prices, thus excluding taxes on products and subsidies..Valuation of output and value added included all taxes on products and subsidies, net of invoiced value added tax.
The 1993 SNA classifies output depending on whether production is traded or not in the market. Market output is output sold at prices that economically significant or otherwise disposed of on the market. Output produced for own final use are goods and services that are retained for the own final use by the owners of the enterprises in which they were produced. Other non-market outputare goods and services produced by non-profit institutions serving households or gvts. that are supplied free, or at prices that are not economically significant.The1993 SNA distinction between market producers, producers for own final use and other non-market producers replaced the distinction in the 1968 SNA between “industries” and “other producers.”
The 1993 SNA recommends the inclusion of work-in-progress in the measurement of agricultural, forestry and livestock production. Work in progress is valued at producer costs plus a notional profit markup. On the demand side, it is allocated either as inventories or gross capital formation depending on the type of product.The 1968 SNA registered agricultural, forestry and livestock production only at the time of the harvest, felling, and/or commercialization. Producers for own final use which were included in the 1968 SNA under “industries” are distinguished as a separate category in 1993 SNA.
The 1993 SNA introduced new concepts in the compilation of the primary distribution and use of income accounts. Mixed income is defined as the operating surplus generated by unincorporated enterprises owned by households. Household actual final consumption is the consumption of goods and services by individual households by expenditure or through social transfers in kind received from the government or non-profit institutions serving households.The 1968 SNA had a wider concept of final consumption expenditure of households than the 1993 SNA, as it included not only what households actually paid, but also health and other expenditures paid or reimbursed by government for services that households are free to select or not. The 1968 SNA did not include the concept of mixed income.
The 1993 SNA makes it clear that the illegality of a productive activity or transaction is not a reason for excluding it from the System. Illegal production could be: (i) production of goods or services that is forbidden by law and (ii) activities which are usually legal but which become illegal when carried out by unauthorized producers. Concealed production and the underground economy are also addressed by the 1993 SNA.The 1968 SNA did not give clear guidance on the coverage of illegal activities in the system of national accounts.
The 1993 SNA broadens the definition of gross fixed capital formationto include expenditure on (i) mineral exploration, (ii) computer software and entertainment, (iii) literary or artisticoriginals, and (iv) military expenses in fixed assets that could be used continuously and repeatedly in production (e.g., airports, roads, docks, other buildings and dwellings); expenses in the form of destructive weapons are excluded fromthe definition of fixed assets.

Like the 1968 SNA, the 1993 SNA continuesto treat expenditures on research and development as intermediate consumption, not gross fixed capital formation.
Expenditures on mineral exploration were treated as intermediate consumption. All purchased software other than pre-installed software supplied with hardware was registered as intermediate consumption. Pre-installed software supplied with hardware was considered capital formation (machinery). The production of entertainment, literary and artisticoriginals was not considered as production. Almost all military expenditures, except those on construction or alteration of family dwellings for personnel of the armed forces, were excluded from gross fixed capital formation.
The 1993 SNA calculates the output of financial intermediation services indirectly measured (FISIM) as the total property income receivable by financial intermediaries minus their total interest payable, excluding any property income receivable from the investment of their own funds. In principle, the total output of FISIM is to be allocated between users—who could be lenders or borrowers— treating the allocated amounts either intermediate consumption of by enterprises or as final consumption or exports. The 1993 SNA recognizes that, in practice, it may be difficult to find a method of allocating the total output among different users; some flexibility in implementation is therefore accepted.The 1968 SNA used the term “bank services charges” when referring to FISIM. The 1968 SNAand 1993 SNA estimate the FISIM concept in the same manner. However, under the 1968 SNA, the whole value of “bank service charges” was allocated to intermediate consumption of a notional industry.
Under the 1993 SNA, the measurement of financial intermediation services provided by central banks is sought along the same lines as FISIM. Central banks’ financial intermediation output is to be valued as the difference between total income receivable and total interest payable, excluding any income receivable from the management of the country’s international reserves. The 1993 SNA recommends flexibility with the implementation of this recommendation, as the estimated output could be small or even negative for some cases. In those circumstances, a solution could be to assess central banks’ output by their operational costs (see SNA News & Notes No3, Jan. 1996).The1968 SNA measured financial intermediation services provided by central banks along the same lines of “bank service charges” applicable to other financial intermediaries.
The measurement of insurance services includes the income accruing from the investment of insurance companies’ technical reserves in the calculation of the services provided to the policyholders.The output of non-life insurance services did not include revenues from corporations’ investments of insurance technical reserves.
The 1993 SNA recommends the valuation of imports of goods on fob basis in the Rest of the World Account.Imports of merchandise were recorded on cifbasis in the Rest of the World Account.
Annex III: Individual Country Analysis

Benin

According to TAIMS, there have been three TA missions on national accounts to Benin: July 2008, February and April 2009.

The number of national accounts mission tasks on data accuracy and reliability doubled between July 2008 and April 2009. Most of the tasks referred to checking the quality and consistency of source data collected by the authorities.

There are issues in terms of methodological soundness that may affect national accounts data concepts and definitions, although the institutional prerequisites of data quality (i.e., coordination and cooperation among source data compilers) are being addressed by the authorities, following the February 2009 mission.

Benin: AFRITAC West Mission Tasks, July 08-April 09
DQAF\Mission datesJul-08Feb-09Apr-09
Number of Mission Tasks
0. Prerequisite of quality070
1. Assurance of integrity000
2. Methodological soundness001
3. Accuracy & reliability306
4. Serviceability000
5. Accessibility000
Total mission tasks377
Percentage Distribution of Mission Tasks
0. Prerequisite of quality0.0100.00.0
1. Assurance of integrity0.00.00.0
2. Methodological soundness0.00.014.3
3. Accuracy & reliability100.00.085.7
4. Serviceability0.00.00.0
5. Accessibility0.00.00.0
Total mission tasks100.0100.0100.0
Source: TAIMS databank and staff estimates.
Source: TAIMS databank and staff estimates.

Burkina Faso

There is only one national accounts mission recorded in TAIMS.

Mission tasks focused on data accessibility (i.e., dissemination) and institutional prerequisites of data quality (i.e., coordination among source data compilers, and coordination between local authorities and donors on the funding of the statistical development plan).

Burkina Faso: AFRITAC West Mission Tasks, Sept. 06
DQAF\Mission datesSep-06
Number of Mission Tasks
0. Prerequisite of quality3
1. Assurance of integrity0
2. Methodological soundness0
3. Accuracy & reliability1
4. Serviceability1
5. Accessibility3
Total mission tasks8
Percentage Distribution of Mission Tasks
0. Prerequisite of quality37.5
1. Assurance of integrity0.0
2. Methodological soundness0.0
3. Accuracy & reliability12.5
4. Serviceability12.5
5. Accessibility37.5
Total mission tasks100.0
Source: TAIMS databank and staff estimates.
Source: TAIMS databank and staff estimates.

Côte d’ Ivoire

According to TAIMS, there have been two TA missions on national accounts to CIV: August 2007 and April 2009.

Mission tasks emphasized methodological soundness. There are a number of issues with the implementation of the data consolidation software (ERETES) and the implementation of compilation methodologies for households’ and government consumption source data, as well as with pending work to change the national accounts’ benchmark year along best international practices.

Data accuracy and reliability were also addressed during missions, mainly referring to statistical techniques used to estimate the GDP deflator and the consumer price index (CPI).

Cote d’Ivoire: AFRITAC West Mission Tasks, Aug. 07-April 09
DQAF\Mission datesAug-07Apr-09
Number of Mission Tasks
0. Prerequisite of quality00
1. Assurance of integrity00
2. Methodological soundness14
3. Accuracy & reliability11
4. Serviceability00
5. Accessibility10
Total mission tasks35
Percentage Distribution of Mission Tasks
0. Prerequisite of quality0.00.0
1. Assurance of integrity0.00.0
2. Methodological soundness33.380.0
3. Accuracy & reliability33.320.0
4. Serviceability0.00.0
5. Accessibility33.30.0
Total mission tasks100.0100.0
Source: TAIMS databank and staff estimates.
Source: TAIMS databank and staff estimates.

Guinea

Since mid-2006, a total of 5 national accounts missions have focused on methodological soundness and data accuracy and reliability. The challenge has been to consolidate the ongoing work on the estimation of GDP at current prices and in volume terms, particularly using standard techniques such as supply and use tables and the sequence of accounts.

Guinea: AFRITAC West Mission Tasks, May 06- July 08
DQAF\Mission datesMay-06Sep-06Jun-07Jul-07Jul-08
Number of Mission Tasks
0. Prerequisite of quality00000
1. Assurance of integrity00000
2. Methodological soundness11103
3. Accuracy & reliability23241
4. Serviceability00000
5. Accessibility10000
Total mission tasks44344
Percentage Distribution of Mission Tasks
0. Prerequisite of quality0.00.00.00.00.0
1. Assurance of integrity0.00.00.00.00.0
2. Methodological soundness25.025.033.30.075.0
3. Accuracy & reliability50.075.066.7100.025.0
4. Serviceability0.00.00.00.00.0
5. Accessibility25.00.00.00.00.0
Total mission tasks100.0100.0100.0100.0100.0
Source: TAIMS databank and staff estimates.
Source: TAIMS databank and staff estimates.

Guinea-Bissau

According to TAIMS, there were two national accounts missions to Guinea-Bissau: August 2006 and May 2009.

Mission tasks emphasized data accuracy and reliability issues (i.e., source data aspects, including the processing of household’s survey and enterprises’ census data, and output data on the mining sector), and data accessibility issues (i.e., documentation of sources and compilation methods). Missions also addressed data serviceability problems related to timely data dissemination.

Guinea Bissau: AFRITAC West Mission Tasks
DQAF\Mission datesAug-06May-09
Number of Mission Tasks
0. Prerequisite of quality31
1. Assurance of integrity00
2. Methodological soundness00
3. Accuracy & reliability46
4. Serviceability01
5. Accessibility23
Total mission tasks911
Percentage Distribution of Mission Tasks
0. Prerequisite of quality33.39.1
1. Assurance of integrity0.00.0
2. Methodological soundness0.00.0
3. Accuracy & reliability44.454.5
4. Serviceability0.09.1
5. Accessibility22.227.3
Total mission tasks100.0100.0
Source: TAIMS databank and staff estimates.
Source: TAIMS databank and staff estimates.

Niger

Since 2006, national accounts missions have extensively advised on source data issues (data collection and processing for different economic sectors) to improve the accuracy and reliability of the national accounts series.

Mission tasks on data methodological soundless addressed the computation of relevant price and volume indices.

Niger: AFRITAC West Mission Tasks, June 06 - April 09
DQAF\Mission datesJun-06May-08Jan-09Apr-09
Number of Mission Tasks
0. Prerequisite of quality1000
1. Assurance of integrity0000
2. Methodological soundness3012
3. Accuracy & reliability0353
4. Serviceability1010
5. Accessibility1000
Total mission tasks6375
Percentage Distribution of Mission Tasks
0. Prerequisite of quality16.70.00.00.0
1. Assurance of integrity0.00.00.00.0
2. Methodological soundness50.00.014.340.0
3. Accuracy & reliability0.0100.071.460.0
4. Serviceability16.70.014.30.0
5. Accessibility16.70.00.00.0
Total mission tasks100.0100.0100.0100.0
Source: TAIMS databank and staff estimates.
Source: TAIMS databank and staff estimates.

Mali

Along with other WAEMU member countries, Mali is embarked in the upgrading of its national accounts series along the 1993 System of National Accounts (1993 SNA) and the implementation of the compilation software ERETES.

Mission tasks focused on data accuracy and reliability, and methodological soundness. Missions reviewed source data collected by the authorities and the statistical techniques for consolidating those data along best international practices.

Mali: AFRITAC West Mission Tasks, June 08-Sep. 08
DQAF\Mission datesJun-08Sep-08
Number of Mission Tasks
0. Prerequisite of quality00
1. Assurance of integrity00
2. Methodological soundness10
3. Accuracy & reliability32
4. Serviceability00
5. Accessibility00
Total mission tasks42
Percentage Distribution of Mission Tasks
0. Prerequisite of quality0.00.0
1. Assurance of integrity0.00.0
2. Methodological soundness25.00.0
3. Accuracy & reliability75.0100.0
4. Serviceability0.00.0
5. Accessibility0.00.0
Total mission tasks100.0100.0
Source: TAIMS databank and staff estimates.
Source: TAIMS databank and staff estimates.

Senegal

National accounts missions to Senegal have been generally targeted to advanced compilation issues recommended by the 1993 SNA (e.g., compilation of financial accounts; development of quarterly national accounts series). Mission tasks focused on methodological soundness, accuracy and reliability, and data accessibility.

Senegal: AFRITAC West Mission Tasks, March 07- April 09
DQAF\Mission datesMar-07May-08Oct-08Apr-09
Number of Mission Tasks
0. Prerequisite of quality3000
1. Assurance of integrity0000
2. Methodological soundness0222
3. Accuracy & reliability2111
4. Serviceability0000
5. Accessibility0111
Total mission tasks5444
Percentage Distribution of Mission Tasks
0. Prerequisite of quality60.00.00.00.0
1. Assurance of integrity0.00.00.00.0
2. Methodological soundness0.050.050.050.0
3. Accuracy & reliability40.025.025.025.0
4. Serviceability0.00.00.00.0
5. Accessibility0.025.025.025.0
0.00.00.00.0
Total mission tasks100.0100.0100.0100.0
Source: TAIMS databank and staff estimates.
Source: TAIMS databank and staff estimates.

Togo

There is only one national accounts mission recorded in TAIMS for Togo.

Mission tasks focused on data accuracy and reliability. Outstanding issues relate to the implementation of the ERETES national accounts consolidation framework, and the establishment of validation techniques to assess the robustness of source data.

Advice on prerequisites for data quality included devising a strategy to implement the methodological recommendations embodied in the 1993 SNA.

Togo: AFRITAC West Mission Tasks, March 09
DQAF\Mission datesMar-09
Number of Mission Tasks
0. Prerequisite of quality1
1. Assurance of integrity0
2. Methodological soundness0
3. Accuracy & reliability2
4. Serviceability0
5. Accessibility0
Total mission tasks3
Percentage Distribution of Mission Tasks
0. Prerequisite of quality33.3
1. Assurance of integrity0.0
2. Methodological soundness0.0
3. Accuracy & reliability66.7
4. Serviceability0.0
5. Accessibility0.0
Total mission tasks100.0
Source: TAIMS databank and staff estimates.
Source: TAIMS databank and staff estimates.
Annex IV. AFW: National Accounts Missions—Program Framework Summary Tasks/Outcomes & Implementation Deadline

(task completed = +, ongoing = + -; number of missions on activity; expected implementation date)

I. Institution and Internal Policy: Includes policy formulation, defining focus points, communication policy, policy meetings, policy evaluation

DQAF

Code
Task & OutcomeBeninBurkina

Faso
Côte

d’Ivoire
GuineaGuinea

Bissau
NigerMaliSenegal
0.4.3Assist the authorities to prepare & implement a technical assistance action plan for the medium-term production of national accounts (human & financial resource availability; a review of trade-offs of key data quality dimensions: timeliness, accuracy/reliability).(+ -; 2;

June

09)
(NA;

NA;

June

2010)
(+ -; 1;

April

2010)
0.4.3Train staff on QNA compilation(+; 1;

Dec. 09)
0.4.3Support the preparation of work program to implement institutional sector accounts.(+-; 1;

ongoing

)

II. Surveys: Includes survey design, questionnaire design, sample design, data collection, data processing, data analysis, data production

DQAF

Code
Task & OutcomeBeninBurkina

Faso
Côte

d’Ivoire
GuineaGuinea

Bissau
NigerMaliSenegal
3.1.1Assist the authorities in developing sound survey questionnaires (i.e., subject to field/pilot testing; observation studies are conducted during the design of questionnaires; data collected are sufficiently detailed to derive national accounts aggregates; sample design ensures that population in scope is represented properly). Set stage for surveys to be conducted on a regular basis.(+, 1,

Jan. 09)
3.2.1Ensure accuracy government finance statistics, merchandise trade statistics, volume and price statistics, and other secondary sources used to compile national accounts statistics is routinely assessed.(+ -;

NA;

NA)
3.3.2Ensure that statistical procedures used to incorporate unobserved activities (including informal, hidden, and illegal activities) follow a detailed, case-by-case approach using specific sources that are most closely related to the estimated variables and pertinent to the reference period.(+ -; 1;

Dec. 09)
3.1.1Source data surveys: Ensure that a comprehensive and up-to-date register provides the basis for sample surveys of mineral, oil & gas enterprises.(+ -; 1;

June

09)
(+ -; 1;

April

2011)
(+ -;

NA;

NA)
3.3.1Statistical techniques: Ensure that data compilation employs sound statistical techniques to deal with data sources (e.g., data compilation procedures (e.g., coding, tabulation) are sound; robust estimation techniques are used to adjust data for missing observations; adjustments for undercoverage follow appropriate guidelines).(NA;

NA;

Dec. 09)
3.2.1Ensure that source data on agriculture are (routinely) assessed for coverage, sample error, response error, and non-sampling errors such as biases, over/under-coverage, misclassification, processing, and nonresponse.(NA;

NA;

Dec. 09)
(+ -; 1;

May 09)
3.4.1Statistical techniques& data validation: Ensure that data compilation of customs trade data (exports and imports) employs sound statistical techniques to deal with data sources (e.g., data compilation procedures (e.g., coding, tabulation) are sound; robust estimation techniques are used to adjust data for missing observations; adjustments for undercoverage follow appropriate guidelines).(NA;

NA;

Dec. 09)
(+ -; 1;

June

09)
(+ -; 1;

July 09)
(+ -; 1;

NA)
(+; 1;

July 09)
3.3.1Use survey data to derive independent (i.e., not as a residual) estimates of household final consumption expenditure.(NA;

NA;

Dec. 09)
(+ -; 1;

Dec.

09)
(+ -, 1,

June

09)
3.3.1Ensure that source data for compiling gross fixed capital formation and services sectors–especially value added of financial institutions–are assessed for coverage, sample error, response error, and non-sampling errors such as biases, over/under-coverage, misclassification, processing, and nonresponse.(+ -; 1;

Dec. 09)
(+ -; 1;

April 09)
3.3.1Statistical techniques: Ensurethat data compilation of the industrial production index employs sound statistical techniques to deal with data sources (e.g., data compilation procedures (e.g., coding, tabulation) are sound; robust estimation techniques are used to adjust data for missing observations; adjustments for undercoverage follow appropriate guidelines).(+-; 1;

June

2010)
(+-; 1,

Jan.

09)

III. Integration Frameworks: Includes design of approach, data collection, checks and corrections, completion, harmonization analysis, integration, and data production

GDP by production

Value added by activity@current prices

DQAF

Code
Task & OutcomeBeninBurkina

Faso
Côte

d’Ivoire
GuineaGuinea

Bissau
NigerMaliSenegal
2.1.1Statistical procedures (e.g., ERETES) & data classification: Ensure that sound adjustments are used to make source data consistent with national accounts requirements in terms of concepts, definitions, classification.(+ -; 2;

Dec. 09)
(+; 1;

mid- 09)
(+ -; 1;

Dec.

2010)
(+, 2,

Dec.

2010)
(NA;

NA;

Dec.

08)
(+; 1;

May 09)
3.1.1Assess whether monthly/quarterly source data are obtained from regular and comprehensive data collection programs that take into account country-specific conditions. Verify if subannual surveys of establishments/enterprises are conducted to obtain detailed quarterly/monthly indicators, consistent with annual data, for most important industrial groups (e.g., at the ISIC one-digit level). Assess if monthly/quarterly data and indicators provide a good basis for compiling expenditure components of GDP.(+ -; 2;

Dec.

09)
(+ -; 1;

June

2009)
(+ -; 2;

Dec.

09)
3.1.1;

3.3.1
• Ensure that the compilation procedures on consumption of fixed capital are sound.

• Ensure that the perpetual inventory method is used as the conceptual basis for estimating consumption of fixed capital.
(+ -; 1;

July

09)
3.2.1Ensure that the accuracy of the information from government finance statistics is routinely assessed.(+ -; 1;

Dec. 09)
(+ -;1;

June

2009)
(+ -; 1;

July

09)
3.3.2Ensure that sound adjustments are employed to make source data on taxes and subsidies consistent with national accounts requirements.(+ -; 1;

Dec. 09)
(+ -; 1;

July

09)
3.1.1Assess if monthly/quarterly data are adequate for compiling reliable quarterly GDP.

• Subannual surveys of establishments/enterprises are conducted to obtain detailed quarterly/monthly indicators, consistent with annual data, for most important industrial groups (& financial corp.).

• Monthly/quarterly data and indicators provide a good basis for compiling expenditure components of GDP.
(+; 1;

mid-09)
3.1.3Timeliness of source data: Assess if the data collection programs provide for the timely receipt of data.(+, 1,

June

09)
4.2.2Ensure that the statistical series is consistent over time.

• Consistent time series data are available for an adequate period of time (at least five years).

• When changes in source data, methodology, and statistical techniques are introduced, historical series are reconstructed as far back as reasonably possible.

• Detailed methodological notes identify and explain the main breaks and discontinuities in time series, their causes, as well as adjustments made to maintain consistency over time.
(+ -; 2;

June

09)
3.3.2Ensure that sound adjustments are employed to make source data consistent with QNA national accounts(+; 1; II

half 09)
GDP by expenditure
Final uses of goods and services @ current prices
3.1.1• Ensure that source data are collected from comprehensive data collection programs that take into account country-specific conditions.

• Annual enterprise/establishment statistics are collected through a regular survey program.

• Household surveys are conducted on a regular basis.

• Comprehensive government finance statistics are available regularly.
(+; 1;

Dec.

09)
(+; 1;

April 09)
3.1.3Timeliness of source data: Assess if the data collection programs provide for the timely receipt of data.(+-; 1;

Sept.

09)
Volume of final uses of goods and services
3.3.2Ensure that sound adjustments are employed to make source data consistent with national accounts requirements (bridge tables between household consumption survey and NA; government financial accounts).(+ -; 1;

Dec.

09)
(+ -; 1;

June

09)
(+ -; 1;

June

09)
3.1.1Ensure that source data for the compilation of unit value indices of exports and imports of goods are collected from comprehensive data collection programs that take into account country-specific conditions(NA;

NA;

Dec.

09)
(NA; NA;

April 09)
3.3.2Provision of TA on the development of SUT.NA; 1,

NA)
(+; 1;

NA)
(+ -; 1;

NA)

IV. Dissemination of Outputs: Includes databases, websites, CD-ROM, printed material

DQAF

Code
Task & OutcomeBeninBurkina

Faso
Côte

d’Ivoire
Guine

a
Guinea

Bissau
NigerMaliSenegal
4.3.1;

4.3.2;

4.3.3
• Ensure that data revisions follow a regular and publicized procedure.

• Ensure that preliminary and/or revised data are clearly identified.

• Ensure that studies and analyses of revisions are made public.
(NA;

NA;

Oct.

09)
(NA; NA;

Sept. 09)
(NA;

NA;

June

2010)
(+; 1; II

half 09)
(+ -; NA;

Dec.

09)
(NA;

NA;

June

09)
(+ -; 1;

Dec.

09)
5.1.2Assist the authorities in preparing documentation of the GDP methodology.(NA; NA;

June 09)
(+ -; NA;

Oct. 09)
5.2Review GDDS metadata, if available.(+ -; 1;

NA)
1

This paper incorporates comments from Tsidi Tsikata, colleagues stationed at AFRITAC East, West and Central, and those at the IMF Statistics Department (STA), including Kim Zieschang and Lisbeth Rivas. Thanks are also due to G. Ukpong from the IMF Executive Director Office and Alan Warburton from the Fund’s Office of Technical Assistance Management.

2

IMF country teams refer to teams from the IMF African department.

4

Other instruments have included training on statistical issues and single or multi-topic technical assistance missions to IMF member countries.

5

For all their wealth of detail, the information from data ROSCs should be used with the understanding that these reports are updated at moderate frequency and may not reflect the most current status of statistical development.

6

Users’ surveys are only included in 6 of the 17 Data ROSCs reports (including updates) written on 13 African countries to date. African countries that have undertaken Data ROSCs include: Botswana, Burkina Faso, The Gambia, Kenya, Mauritius, Mozambique, Namibia, Niger, Senegal, South Africa, Tanzania, Uganda, and Zambia.

7

AFRISTAT, in cooperation with AFW, is developing a series of indicators and macroeconomic ratios that allow both to assess the reliability and analyze the quality of national accounts data. The related documentation is expected to be presented to AFRISTAT member countries, including AFW member countries, in late 2009.

8

See, for example, Roberto Olinto Ramos et. al. (2008), op. cit.

9

Areas of action include, among others: launching of economic and population censuses; fielding household income and expenditure budget surveys, updating of business registries; improving survey methods of financial and nonfinancial public corporations; and developing “satellite accounts” for main export sectors.

10

See Annex II for a summary presentation of main methodological differences between the 1968 SNA and the 1993 SNA.

11

SDDS was established by the IMF in 1996 for member countries that have or might seek access to international capital markets, to guide them in providing their economic and financial data to the public. SDDS requires the development of a national summary data webpage, an advanced/early data release calendar, and the production of high-frequency data for 18 major macroeconomic time series, including quarterly national accounts.

12

The need to address data dissemination issues is in line with recent views shared by technical assistance providers in national accounts working in the region (mainly: AFW, AFRISTAT and the INSEE-France). According to AFW staff, this data quality aspect is being stressed in recent and prospective technical assistance missions, although it is recognized that this is a sometimes difficult step for managements of statistical offices to take, since it can expose their organizations to public criticism of the quality of the data they are releasing, and of the quality of management of the office. On the other hand, laying out the data along with its weaknesses through the web can be an effective first step toward attracting national and international resources to under-resourced data producing institutions.

13

Guinea recently updated its national accounts base year to 2003. However, the information received from the IMF country team suggests that incidence of data inconsistencies, notwithstanding reported improvements in national accounts’ data sources and methodology.

14

Sources: Olinto-Ramos et. al. (2008) “Latin America: Highlights from the Implementation of the System of National Accounts 1993 (1993 SNA), IMF Working Paper, WP/08/239, October

Other Resources Citing This Publication