IMF Policy Paper: Report On The Technical Assistance Evaluation Mission To Transition Economies: Albania And Georgia

This evaluation of technical assistance (TA) and training in statistics looks at the experience of two transition economies, Albania and Georgia, during roughly the period 2005–2010. The TA, including the training, to these countries covered all the topical areas on which the IMF’s Statistics Department’s (STA) focuses, i.e., national accounts, price statistics, and monetary, balance of payments and government finance statistics, albeit with differing emphases between the two countries. Part of the assistance was funded directly from the IMF’s budget, while other elements (in particular the peripatetic advisors) were financed externally, in these cases by the Japanese government.

Abstract

This evaluation of technical assistance (TA) and training in statistics looks at the experience of two transition economies, Albania and Georgia, during roughly the period 2005–2010. The TA, including the training, to these countries covered all the topical areas on which the IMF’s Statistics Department’s (STA) focuses, i.e., national accounts, price statistics, and monetary, balance of payments and government finance statistics, albeit with differing emphases between the two countries. Part of the assistance was funded directly from the IMF’s budget, while other elements (in particular the peripatetic advisors) were financed externally, in these cases by the Japanese government.

I. Introduction

1. The evaluation of technical assistance (TA), including training, to member countries provides an opportunity for both TA providers and recipients to draw lessons that can inform efforts to enhance the effectiveness of TA programs. Following the recommendations of the IMF’s Independent Evaluation Office in 2005, which called for more systematic approaches to assess progress on major TA activities and to identify factors contributing to success and/or shortfalls of TA delivery and implementation, such evaluations are becoming an integral component of IMF TA activities. For countries that receive TA in macroeconomic statistics, where the dissemination of data to the public is critical for better decision making, evaluation missions also offer a chance to discuss with data users and help to enhance the dialogue with data compilers to inform statistics processes and better products.

2. Accordingly, staff of the IMF’s Statistics Department (STA) visited Albania and Georgia to evaluate the department’s TA programs in two transition countries over roughly the period 2005 to mid-2010.1 Both countries had received a sizable amount of TA and training, with TA comprising traditional short (usually two-week) TA missions as well as regional or peripatetic advisors, while training involved largely STA’s courses in the various topical methodologies, delivered out-of-country in the IMF’s regional training centers or at headquarters in Washington.

3. In both countries the mission visited the statistics agency, the central bank, and the Ministry of Finance (MoF), and also met with users. The evaluation is based on the missions’ discussions and was also informed by a review of relevant documents. The evaluations contained three elements. First, a survey was conducted in advance of the visits at each of the TA-receiving institutions using a standardized questionnaire to solicit views on various aspects of STA’s TA program. Each survey was accompanied by a detailed matrix to report on the status of implementation of the action plans for the respective area. Second, an IMF desk review was conducted to ascertain each country’s data provision to STA and the information available at headquarters on the implementation of missions’ recommendations in each topical area. Third, discussions were conducted with key representatives of the agencies receiving TA to seek clarification on the main issues arising from the survey responses and desk reviews, as well as to solicit views on key aspects of program design and implementation. Discussions were also undertaken with representatives of the major user groups including government, private research groups, and bankers. In both countries, presentations were made on the last day of the mission to present the mission’s findings to representatives from the compiling agencies.

4. In both countries the Data ROSC was used as a benchmark from which to assess progress. For Georgia the Data ROSC was conducted in 2002, for Albania in 2006. The Data ROSC provided a snapshot of the state of statistics at the time, as well as a series of recommendations for statistical improvement.

5. The report is structured as follows. Section II contains a brief overview of STA’s TA program to Albania and Georgia. Section III reviews STA’s TA for Albania. Section IV reviews STA’s TA for Georgia. Section V presents the conclusion of the evaluation and recommendations. A number of appendices complete the report.

6. The mission would like to express its appreciation to the statistics-compiling agencies of Albania and Georgia for the excellent hospitality, cooperation, and assistance extended to the mission. The enthusiasm and dedication of the mission’s counterparts greatly facilitated the mission’s work.

II. Brief Overview of STA’ s TA Program to Albania and Georgia

7. Albania started participating in the General Data Dissemination System (GDDS) in 2000, and Georgia in 2006. Both have used the GDDS metadata disseminated on the Dissemination Standards Bulletin Board (DSBB) and national websites to coordinate their

TA.

8. On May 17, 2010 Georgia subscribed to the SDDS. Albania too seems to be relatively close to being able to subscribe to the SDDS, although more work remains to be done in a few areas.

9. Georgia has received a sizable amount of TA from STA in most statistical topical areas. In conducting this TA, STA has worked closely with the National Bank of Georgia (NBG), the MoF, and the National Statistics Office of Georgia (the forerunner of GEOSTAT—see below). The relative importance of statistics in the IMF’s overall TA program to Georgia is demonstrated by the assignment of STA’s country coordinator to coordinate the preparation of technical assistance country strategy notes (TACSN).

10. In 2002, STA conducted the Report on the Observance of Standards and Codes—Data Module (Data ROSC) mission to Georgia to assess the overall quality of the macroeconomic statistical system, and provided extensive recommendations for its further development. Its summary conclusion was that statistics in all the macroeconomic datasets were reasonably adequate for surveillance and other purposes, but external statistics were not. The Data ROSC recommendations, as well as Georgia’s objective of achieving SDDS participation, provided a structured approach to the TA delivery to the country. Technical assistance has consequently been provided in national accounts (two missions, in 2004 and 2005), price statistics (three missions, in 2005, 2006, and 2007), balance of payments (BOP) and international investment position (three missions, in 2003, 2005, and 2006), monetary and financial statistics (in 2009), government finance statistics (in 2006), and in the Fund’s data initiatives (two missions, in 2006 and 2009). In addition, reflecting the relative weakness in external statistics identified in the Data ROSC, from November 2006 to November 2008, a long-term regional statistical advisor in external sector statistics based in a neighboring country assisted Georgia with six peripatetic two-week missions.

11. Over the 2005-2009 period, 30 places were granted to compilers from Georgia in various STA training courses and seminars. These included participation in general training activities at the IMF Institute (INS) in Washington, D.C. and at the Joint Vienna Institute (JVI), as well as short topical seminars tailored to the specific group of countries (e.g., International Trade in Services Seminar conducted by the Interagency Task Force on Statistics of International Trade in Services (TFITS) with participation of the IMF staff).

12. Albania’s Data ROSC was undertaken in 2006, and provided the framework for much of the TA in the evaluation period. The TA program comprised TA missions covering all the major topics in macroeconomics. In recent years, technical assistance has been provided in national accounts (two missions, in 2005 and 2009, and four peripatetic visits by a statistics advisor during 2008), price statistics (one mission in 2007), balance of payments and international investment position (two missions, in 2006 and 2007), government finance statistics (in 2008), and in the Fund’s data initiatives (two missions, in 2006 and 2010). TA recipient institutions in Albania were the Bank of Albania (BoA), the MoF, and the Institute of Statistics (INSTAT). Most of STA’s TA to Albania was delivered in the form of short-term staff missions, except for the TA for national accounts statistics, for which the TA program included a series of three-week peripatetic expert visits.

13. Over the 2005-2009 period 19 INSTAT, BoA, and MoF staff members attended STA training courses. Compilers have been given access to both general training activities at the INS and JVI, and to STA’s special-topic regional training seminars (in Ljubljana, Slovenia).

Albania and Georgia: Key Capacity Building Achievements During 2005-2010

article image

STA’s Technical Assistance Evaluation Missions to Albania and Georgia: Key Achievements in Developing Statistical Products and Assessment of Risks

article image
article image

III. Albania: TA Evaluation by Recipient Agency

14. The mission held evaluation meetings with senior officials and technical counterparts of all TA-recipient agencies, including the BoA, the MoF, and INSTAT.

The following subsections document the main discussions with these agencies. The mission also reviewed the completed questionnaires by these agencies, a summary of which is included in Appendix Table 3.

A. SDDS Initiatives in Albania

15. Albania has been making continued progress towards SDDS subscription, but

road blocks remain, and the country has been overtaken by others (Appendix Table 4).

16. The weak institutional framework, the lack of adequate resources, and the limited public awareness of the SDDS continue to hamper preparations for SDDS subscription. The mission discussed with the various agencies the need to improve interinstitutional communication and to document and update progress, while prioritizing the statistics program action plan with deliverables and timetables, as a major imperative to success. Analysis of the institutional framework and statistical practices undertaken in each topical area revealed that the level of observance of the SDDS requirements varies across the three institutions. While the data disseminated by the BoA and MoF already are in line with SDDS requirements (with the exception of IIP statistics), several categories of data disseminated by INSTAT lack sufficient frequency, timeliness and/or coverage. Nevertheless, there have been a number of significant recent improvements. These include (1) developing SDDS metadata for all data categories; (2) updating the National Summary Data Page (NSDP); (3) developing an Advance Release Calendar (ARC), (4) disseminating quarterly general government operations data, and (5) adopting and disseminating a policy on data revisions.

17. More work remains to be done for a number of data categories. These include: (1) current-price quarterly GDP data expressed in national currency remain to be disseminated; (2) the coverage, accuracy, and timeliness of the FDI survey conducted by INSTAT needs to be improved to meet SDDS requirements for IIP statistics;

(3) the industrial production index (IPI) needs be disseminated with monthly frequency, and within six weeks of the reference period (as opposed to the current quarterly frequency and issuance within 75 days of the reference period); (4) the producer price index (PPI) to be disseminated monthly within one month of the reference period; and (5) the coverage of various categories of quarterly labor market statistics—employment, unemployment and wages—need to be improved (with support from Eurostat and the International Labour Office (ILO), the labor force survey will be conducted starting from 2011 with the questionnaires and survey frame refined to achieve these improvements). Achieving SDDS subscription would also be much facilitated by enhancing the role of the SDDS correspondents in INSTAT and the MoF, and improving their coordination with the SDDS country coordinator, so as to jointly develop a prioritized action plan. All three institutions would need training in techniques of data dissemination.

18. The TA on data dissemination (the consultation mission in 2006 and the SDDS assessment mission in 2010) was in general received well in all three government agencies. The missions were found to be useful in resolving institutional issues and in developing SDDS metadata. However, a desire was expressed for TA to develop a specific prioritized action plan with a time table to facilitate Albania’s subscription to the SDDS through helping channel TA into critical areas that need to be addressed.

B. Institute of Statistics

19. In recent years, there has been a substantial enhancement of INSTAT’s statistical capacity. In 2008, the NA unit was upgraded into an NA Directorate (NAD) with 11 employees. Following the recommendation of the peripatetic STA TA advisor, the NAD was reorganized into three sections: annual accounts, quarterly accounts, and NA development. In June 2010, the Government demonstrated its commitment to macroeconomic statistics by further enhancing INSTAT’s human resources through providing resources for the hiring of 47 new staff, among them are 30 experts, nine of whom will join the NA Directorate. Staff salaries have been adjusted accordingly. These measures are likely to be helpful in combating staff turnover, which already has been low in recent years (with some exceptions among the Information Technology (IT) staff).

20. The Statistical Board established by the 2004 Statistics Law became operational at the end of 2005. The Statistical Board comprises members from academia, the MoF, and the BoA. It provides significant technical support to INSTAT, and helps to create synergies between INSTAT and wider circles of users, as well as to establish public awareness and a statistical culture. The Program of Official Statistics for 2006-2011 was approved by Parliament on January 31, 2008. Memoranda of Understanding have recently been signed between INSTAT and other data-producing agencies, namely the BoA, the Ministry of Agriculture, the Tax Authorities, and the High Inspectorate of Declaration and Audit of Assets. However, some bilateral institutional issues are still to be resolved, such as the coverage and timeliness of the FDI survey that is compiled by INSTAT for the BoA on a commercial basis and used as a data source for external sector statistics.

21. Despite these significant measures, more remains to be done to strengthen the institutional capacity to produce high-quality statistics, including the further development of the business register, data collection procedures (at present, all business surveys are conducted by interview, as a traditional practice and because the business register does not currently hold postal addresses for enterprises), and the enhancement of statistical techniques (classifications, surveys), staff skills, quality control, and knowledge transfer.

Modalities of TA delivery and major outcomes

22. During the period 2005 to mid-2010, the larger part of STA’s TA program to INSTAT has been delivered through a series of visits from the NA peripatetic advisor in national accounts (4 three-week missions in 2008 and one two-week mission in 2009) preceded by a short three-day staff assessment visit in 2005. In addition, a one-week mission in price statistics was conducted in 2007. The TA was delivered predominantly by an expert, reflecting both the shortage of STA staff and the need for an intensive mode of delivery over a relatively short period of time in complex areas. Albania’s Project Management System (PMS) including expert’s backstopping from STA was the keystone for the TA provided in real sector statistics.

23. INSTAT has also benefited from the extensive TA offered and financed by the

EU. The Twinning Project Albanian Statistics toward EU has been implemented with the technical assistance of the Italian National Institute of Statistics, in partnership with Statistics Finland and the Hungarian Central Statistical Office. The IPA 2007 Multi-beneficiary Statistical Cooperation Program (IPA 2007) in Albania started in April 2010. IPA 2007 provides TA for the implementation of censuses of enterprises (agriculture and nonagriculture) and extensive TA on NA. It is considered that the classifications, statistical units’ model, strategy development for supply-and-use tables (SUT), regional GDP methodology, and institutional sector accounts development be included in the IPA 2007 project. TA in price statistics was formerly offered exclusively by STA. Starting from 2008, Albania has begun a project with SIDA, which comprises two to three one-week price statistics missions each year.

24. INSTAT indicated general satisfaction with the mix of STA’s TA program.

INSTAT managers were very appreciative of the long-term STA expert in NA, both for his expertise and knowledge of local conditions and the very effective hands-on training that he provided. The compilers liked particularly that the expert was working together with the staff directly, rather than receiving data and work from them separately. It was also noted that longer duration (visits of perhaps two months) would work better for TA delivery to INSTAT. There is also a feeling that the TA could have benefited more from a better cooperation between all TA providers.

Real sector statistics

TA delivery and outcome

25. The March 2006 Data ROSC assigned four LNO ratings to the national accounts statistics including on resources, scope (the accounts did not meet the minimum requirement set by the Intersecretariat Working Group on National Accounts); the source data (which were untimely and hampered by poor questionnaire design, as well as limited area coverage of the Household Budget Survey (HBS), insufficient frequency of surveys, and insufficient estimates of the nonobserved economy); and statistics techniques (estimates of production from nonobserved activities were based on limited labor-input data and productivity assumption; estimates of expenditure components relied on unusually weak assumptions, were not cross checked against other data sources, and were based on a limited benchmark).

26. The 2008 STA TA project that followed the Data ROSC prepared a long-term strategic plan for the development of the Albanian system of national accounts.

Improvements have been made in a number of key data sources (the labour force survey and a country-wide HBS have been conducted); the quarterly GDP series for 2002-2008 by production approach was developed in 2008, and has been published each quarter since December 2008; and improvements have been made to the methodology used in the annual national accounts. Changes in some key annual source data resulted in new data becoming available for the 2008 reference year during 2009. During 2009, INSTAT reviewed and revised the Annual Structural Business Survey (ASBS). The structure of the questionnaire has been improved, and additional questions have been included to provide data on changes in inventories and improve the data on gross fixed capital formation. Data for the government’s nonmarket activities became available as of Q1, 2008. However, the implementation of some aspects of the strategic plan has been postponed, waiting until the start of the NA component of IPA 2007, which will also include censuses of agriculture and nonagriculture enterprises. The publication of the reference guide on data sources and methods for real sector statistics Description of sources and methods for GNI for Albania was prepared within the framework of IPA 2007.

27. The TA in price statistics has also followed the finding of the 2006 Data ROSC

that noted outdated CPI weights, unsatisfactory coverage of the HBS, which only included urban households, and shortfalls in verification procedures. The 2007 TA mission recommended a new approach for selecting the list of CPI market basket items, which was adopted and updated thereafter. A new annual national-level HBS is now conducted with data available for both urban and rural households at the 2, 3, 4, 5, and 6 digit Classification of Individual Consumption According to Purpose (COICOP) levels. The basket is renewed accordingly. Each household living in an owner-occupied dwelling or in rent free housing was asked to estimate their rent on the open market, with data to be used directly for weights for the new index.

28. Thus, most of the Project Framework Summary (PFS) action plan items and Data ROSC recommendations for price statistics have been implemented. The exception is the timeliness and periodicity of the PPI, which is delaying SDDS subscription. This would however be difficult to overcome, due to the specifics of the present compilation procedures (data are collected by enumerators). Geometric means are now generally used for elementary aggregates in price indices, especially where weight data are not available for individual price quotations. The price relatives are computed at the individual price quotation level and aggregated to the item level employing a current period to previous period matched price procedure and the geometric mean formula. There is still room for improvement for the construction index and for applying the quality adjustments.

Training for INSTA T staff

29. INSTAT staff attended five training courses in real sector statistics including a training course on NA at the INS (2006), a quarterly NA course at the JVI (two staff, in 2008), and price statistics courses at the INS (2005) and the JVI (2006). The four missions of STA’s 2008 NA project included a significant training component. The 2009 NA mission provided on-the-job training in the relevant concepts of quarterly GDP, including those relevant to developing a quarterly expenditure measure of GDP.

TA needs for the period ahead

30. Significant amounts of assistance in NA will be provided to INSTAT by the EU in the years ahead, but a significant role remains for TA delivery by the IMF, especially in light of its flexibility in delivering targeted assistance. Phase 3 of the EU IPA 2007 for years 2010-2012 is planning to include TA in compiling the quarterly national accounts by expenditure approach; SUTs, balancing commodity accounts, computing systems for final balancing; and developing sector accounts. Phase 4 for 2013-2014 is planning to compile symmetrical input-output tables, quarterly national accounts by income approach, and annual SUTs. Phase 5 for 2014-2016 plans to produce five-yearly symmetrical input-output tables. Future STA TA to INSTAT should be considered carefully in close collaboration with EU.

31. INSTAT requested more access to IMF training for price courses, and longer NA courses on methodology. This increased demand for training in real sector statistics is related to the recent enhancements in staff complements, and would serve to support INSTAT’s reform programs.

C. Bank of Albania

32. The BoA is responsible for compiling monetary and financial statistics, and external sector statistics. Monetary and financial statistics (MFS), BOP and IIP statistics are reported by the BoA to STA, although there are technical difficulties in transmitting data using the integrated correspondence system (ICS). The BoA submits data to the IMF-World Bank Quarterly External Debt Statistics database (QEDS). However, it has not yet embarked on Financial Soundness Indicators (FSIs) data compilation. The BoA has agreed to participate in Coordinated Direct Investment Survey (CDIS), but yet to submit data for year 2009. The BoA coordinates Albania’s SDDS preparations.

33. There has been a significant increase in resources provided to statistics by the BoA in recent years. In 2007, a new IIP unit was added to the external sector statistics section of the Statistics Department. In 2009, three expert positions were added to the Statistics Department, resulting in a total staff of 26, of whom 12 are assigned to external sector statistics. One of the major staff resource concerns of the Statistics Department is the lack of IT experts. Two attempts to hire an IT expert in the last three years were unsuccessful.

Modalities of TA delivery and major outcome

34. The 2006 Data ROSC gave a low rating to the source data for external sector statistics, which were reported to be based on infrequent and/or incomplete surveys, including those on freight, insurance, and smuggling adjustments for imports; those on travelers’ expenditures domestically and abroad; on workers’ remittances; on direct and portfolio investment; and on trade credit. Grants from abroad to regional and local governments were underreported. STA’s 2006 and 2007 TA missions delivered by staff focused on addressing these issues.

35. Progress was achieved in remedying the deficiencies. In cooperation with INSTAT, a quarterly survey on remittances was started in the first quarter of 2006. Following the outcome of the 2006 FDI survey, annual IIP data have started to be compiled, albeit with limited coverage. The IIP statement has been reconciled with the financial account. Improved annual IIP (for 2008 data) and EDS were published in April 2010 for the first time in

time series format on the BoA website, and in the annual report with methodological notes.

36. The further development of data sources seems to be constrained at the moment by the commercial agreement between the BoA and INSTAT under which INSTAT conducts surveys on behalf of the BoA at the expense of the BoA. Any cost increases beyond the budget approved by the BoA create delays and affect the BoA’s statistical production. Moreover, the IT infrastructure and support remain inadequate. The BOP statement is generated through a set of interlinked excel files. Data entry and data verification procedures are manual. IT support has been only available for the implementation of simple data entry programs for the travel and FDI surveys. A project for the implementation of an electronic data reporting system is at its initial stage. The banking reporting system also needs improvement for the BOP compilation.

37. The BoA expressed broad appreciation for STA’s TA in external sector statistics, but suggested improvements, namely that it would be helpful to widen focus of the TA to more than one BOP component and to avoid overlap between missions’ recommendations.

Training for BoA staff

38. BoA staff attended six IMF training courses in external statistics, including two BOP courses in the INS (2005 and 2008), two BOP courses in the JVI (2006 and 2008), one EDS course in the JVI (2008) and one in the INS (2009), as well as two regional seminars in Ljubljana, Slovenia, on CDIS (2008) and Remittances (two participants, 2009). Participants recognized the value of short specialized courses for experienced staff, and longer general courses for the newcomers. Difficulties were noted in the proper channeling of the course solicitation forms so as to ensure the right level of participation: it is, for instance, important for STA to emphasize the statistical nature of the courses for those designed specifically for compilers.

TA needs for the period ahead

39. Under EU auspices, the BoA has been twinned with the Bank of Italy and Bank of France. The twinning project is envisaged to cover: (1) design of the development plan in the areas of BOP/IIP statistics and MFS; (2) delivering TA in compiling stocks and flows in FDI and portfolio investment; (3) staff training on EU/European System of Central Banks (ESCB) output requirements and methodological standards for BOP/IIP statistics (implementation of BPM6 recommendations) and for MFS; (4) “on-the-job” training in data collection methods, the design of questionnaires, business registers, and IT solutions for data collection and information storage, with special attention to the stocks and flows of FDI and portfolio investment; (5) choice of a model for the strengthening of the present BOP/IIP compilation method, and for its gradual approach to EU/ESCB standards and definitions;

(6) the development of the basis for transaction accounts for the banking system; and

(7) the identification of the links between the banking system and the external statistics, and of the appropriate methods for revaluations. A project aimed at the implementation of an electronic reporting system has started with the assistance from the Bank of Italy. It is expected to be completed between the end of 2011 and the middle of 2012.

D. The Ministry of Finance of Albania

40. The Organic Budget Law and the Guidelines on State Budget Performance clearly mandate the responsibility of the MoF for compiling and disseminating government finance statistics (currently on a modified cash basis). The MoF has created a separate section under the General Treasury Directorate for the Financial Reporting which has the responsibility to compile the Fiscal Budget Consolidated Indicators table (monthly/annual) and GFS (annual) and IFS (monthly) tables. This section is supported by the counterpart section within the IT Department. The reporting of the GFS statistics to STA using the ICS has not yet started.

Modalities of TA delivery and major outcome

41. Progress in GFS compilation has been facilitated by a strong dedicated compilation unit and the commitment of the authorities. The 2006 Data ROSC noted the existence of progress in several technical areas at the MoF—including the design and implementation of an accrual-compatible database infrastructure—as well an advanced migration path to the Government Finance Statistics Manual 2001 (GFSM 2001).

A one-week STA expert mission visited the MoF in February 2008. The MoF regards the quality of STA’s TA delivery highly.

42. Starting from 2008, a new government accounting system, AMoFTS, has become partially operational. The system is operated by Oracle E-Business Suite accounting software, on line with 36 Treasury District Offices, and has the facility for electronic payments with the BoA. There are still issues in generating balance-sheet transactions of the Treasury, including all special funds and debt transactions, the solving of which is dependent upon the banking system and the Tax/Customs Administration systems (which are not compatible with AMoFTS system). Also, the AMoFTS is not able to generate modified accrual debt data. The General Treasury Department reports monthly/annual data of external donor financed projects only through the implementation of the project unit’s reports, and does not check this information against the data from the donors’ agreements.

Training for MOF staff

43. MoF officials considered STA’s training activities in GFS very effective. During 2005-2009, five staff from the MoF attended GFS statistics courses at the INS and the JVI, and one person attended the External Debt course at the JVI.

TA needs for the period ahead

44. MoF officials envisage that STA’s TA could contribute in all processes of budget/treasury and debt transactions, specifically to moving to modified accrual basis accounting by assisting in designing and developing reports for compilation of the GFS data from the new AMoFTS system database; and adopting the accrual basis for recording data on government debt.

45. In designing the future GFS TA program, it is worth noting that the MoF is included in the IPA 2007 project, which is planning by 2015 to connect online all general government units. Under this program it is planned also to increase the skills and capacity of the professional staff and to modernize the electronic equipment. STA’s future TA needs to be planned in coordination with this work.

46. The MoF stressed the high-level demand for STA GFS and external debt training. There was some discussion on the feasibility of organizing a regional training course.

IV. Georgia: TA Evaluation by Recipient Agency

47. An STA mission to evaluate STA TA visited Tbilisi during July 16-18, 2010, and

met with NBG Deputy Governor Mestvirishvili, GEOSTAT Executive Director Chelidze, and MoF Financial Policy Department Head Kavelashvili, and their respective staff. The mission also met with selected users, drawn largely from the financial institutions in the country. The following section documents the main discussions with these agencies. The mission also reviewed the completed questionnaires by these agencies, a summary of which is included in Appendix Table 7.

A. SDDS Initiatives in Georgia

48. Georgia’s subscription to the SDDS on May 17, 2010 marked a major step forward in the development of the country’s statistical system, reflecting the authorities’ serious commitment, as well as STA’s TA program in this area. STA’s TA program involved assisting the authorities in addressing the institutional issues related to SDDS subscription, developing the detailed action plan, and facilitating interagency coordination, and helping to populate National Summary Data Pages (NSDPs) with the SDDS data categories in accordance with the electronic procedures established by the IMF to facilitate quarterly certification of the national metadata. GEOSTAT, the NBG, and the MoF made substantial efforts to achieve these objectives. The SDDS TA (two missions, in 2006 and 2010) was highly regarded by all three agencies; the last mission had been seen as critical in identifying remaining blockages and providing guidance on how the agencies could work through them.

B. The National Bank of Georgia

49. The NBG is responsible for compiling monetary and financial statistics, and external sector statistics. MFS, BOP and IIP statistics are reported by the NBG to STA

using the ICS on a timely and regular basis for publication in the International Financial Statistics (IFS) and Balance of Payments Statistics Yearbook (BOPSY). The NBG participates in the IMF’s Coordinated Portfolio Investment Survey (CPIS) and the CDIS, and submits data to the QEDS. However, Georgia does not participate in the coordinated compilation exercise (CCE) for FSIs conducted by the IMF, and has yet to begin compiling FSIs.

Modalities of TA delivery and major outcome

50. TA delivery covered both monetary and external statistics, with heavy focus on

the latter. Between 2005 and mid-2010, STA delivered two TA missions in external sector statistics (in 2005 and 2006), and one mission in MFS (2009). In addition, from November 2006 to November 2008, a long-term regional statistical advisor in external sector statistics (RSA) assisted NBG through six peripatetic missions. NBG staff attended 12 training activities in monetary and external sector statistics, including training courses at the INS and JVI, regional topical seminars conducted by STA on FSIs, CDIS, Remittances, the International Transactions Reporting System (ITRS), as well as the International Trade in Services Seminar conducted by the United Nations Statistics Division (UNSD) under the auspices of the Interagency Task Force on Statistics of International Trade in Services (TFITS) with participation of STA staff. Staff from the NBG received hands-on training in methodology and data compilation of external sector statistics from the RSA.

Monetary statistics and Financial Soundness Indicators

TA delivery and outcome

51. STA is a sole provider of TA in MFS to the NBG. The TA was provided by staff from IMF headquarters.

52. The July 2002 Data ROSC provided recommendations on a framework for further developments in MFS. The ROSC called for improvements in the statistical coverage of nonbank depository corporations, improved transparency, and the provision of documentation on metadata. In accordance with the Monetary and Financial Statistics Manual, 2000 (MFSM), the NBG has begun compiling monetary data using the framework of the STA’s Standardized Report Forms (SRF). These data have been published in the IFS Supplement since September 2006. Historical time series have been revised in accordance with the SRFs back to December 2001. Recently, the publication on the methodology for compiling the MFS was updated, further elaborated, and redisseminated. A special telephone line was created in the NBG to communicate with users.

53. The March 2009 TA mission was provided by STA a significant time after the initial request from the authorities. The 2006 Data ROSC found that the coverage of monetary statistics needed to be expanded, and several classification issues were to be addressed. Following the mission, the authorities began their work on improving the classification of financial instruments and expanding coverage to include microfinance companies and insurance companies. A number of special deposits accounts in national and foreign currencies have been reclassified and included in broad money. The inclusion of the accounts of the credit unions, also recommended by the mission, was not pursued by the NBG because of their statistical insignificance, but the regular monitoring of their operations was set up on the basis of quarterly data on total assets and total liabilities collected by the Supervisory Department.2 To develop data on insurance companies, the Monetary Statistics Division is working in cooperation with the Supervisory Department to insure that data currently collected on a quarterly basis provide a satisfactory basis for compiling the accounts of other financial corporations (OFCs). The chart of accounts has been revised and implemented from 2010 data.

External sector statistics

TA delivery and outcome

54. The July 2002 Data ROSC assessed compilation of the balance of payments statistics as generally inadequate, with largely non-observed ratings for a whole range of data quality elements, starting from legal settings and resources to insufficient source data, poor statistics techniques, general irrelevance of BOP statistics, and inadequate dissemination practices. Following the Data ROSC, STA fielded three TA missions conducted by STA experts (2003, 2005, and 2006). In addition, to fully meet Georgia’s statistical needs, STA’s RSA in external sector statistics, the IMF headquarters staff on special assignment, stationed in Azerbaijan, assisted Georgia via six peripatetic missions during 2006-2008. The authorities identified this mode of TA delivery as being particularly effective. The authorities commented that the missions’ TA reports depicted the situation existing at that moment very precisely, and that all the recommendations and methodological instructions contained in the TA reports were very useful and timely. Georgia’s Project Management System (PMS) was a important component shaping the TA provided in the external sector statistics.

55. The TA program assisted the NBG in a number of important activities targeted at statistical capacity building and reforms in external sector statistics. In 2006, the BOP compilation function was transferred from the State Department for Statistics (now GEOSTAT) to the NBG. The RSA assisted the NBG in refining the legal settings for compiling external sector statistics and improving the process of interagency data sharing. The proposed amendments to the Law on the NBG were ratified by the Parliament on March 14, 2008. The Law clearly assigned primary responsibility to the NBG to compile the external sector stock and flow statistics. To facilitate coordination and data sharing between the NBG and other data-producing agencies, the NBG representative became a member of the GEOSTAT Board, which serves as the National Statistics Council. However, more remains to be done to institutionalize an adequate data sharing between data-producing agencies. Within the NBG, a resolution was signed and put into operation to ensure the smooth flow of the external sector data between the Balance of Payments Statistics Division and other NBG units. The staff resources of the NBG’s BOP Division were augmented. The institutional arrangements were set up to ensure participation of Georgia in the QEDS, CPIS, and CDIS.

56. The source data also were considerably improved. In January 2009, the NBG implemented the ITRS, which became a foundation of the balance of payments compilation. The preparatory work assisted by RSA included: (1) institutional measures implemented by management of the NBG with regard to IT and staff resources to support the ITRS application; (2) resolving legal issues to support the mandate ITRS reporting; (3) development of the IT for all ITRS constituencies; (4) the ITRS design and field testing, (5) elaboration of the ITRS data verification and validation procedures, (6) the ITRS consulting and training for the reporting banks. To ensure sustainability of the ITRS the computer resources are to be improved. The program for compiling the gross external debt statistics was developed and resulted in submission of the QEDS data required by the SDDS requirements. The NBG has launched compilation of the quarterly IIP, which is published in the BOPSY and IFS, starting from November 2007. The NBG finalized Georgia’s Reserve Template according to International Reserves and Foreign Currency Liquidity: Guidelines for Data Template, which was disseminated in May 2010.

57. The TA program aimed at further facilitation of the policy analysis and resolving credibility issues regarding the external sector statistics. Discrepancies were identified between the data used by the IMF’s area department, and those provided by the NBG to STA. An aggregated BOP presentation format was developed and agreed by all parties, and corresponding links were provided to the specific components of the standard and analytical presentations of the BOP in the country’s Economic Data Sharing System. At the end of 2007, the NBG launched a new publication of Georgia’s external statistics starting with the annual data for 2006. The publication contains about 150 pages of stock-and-flow data, charts, tables, analytical notes, and description of sources and methods.

58. The methodology and statistics techniques for recording specific external sector transactions were refined. These include the recording of external debt data from the state-owned enterprises, selected international trade in services data, international reserves, remittances, transactions related to the Georgian segment of the Baku-Tbilisi-Ceyhan pipeline, transfer pricing, and other transactions related to the oil and gas sector.

59. The TA program aimed also at improved regional cooperation. In 2007, the RSA facilitated a visit of the BOP compilers from the National Bank of Azerbaijan to the NBG. In 2008, compilers from the NBG visited the National Bank of Ukraine to benefit from the ITRS experience in Ukraine. In 2009, compilers from the National Bank of Armenia visited the NBG to learn about the ITRS implementation process.

Training for NBG staff

60. The two training courses on MFS (2005 and 2007) and one training seminar on FSI (2007) held at the JVI were well received by the MFS compilers. In addition to the on-the-job training during the TA missions, NBG staff attended one BOP course at the INS (2007), two BOP courses in the JVI (2008 and 2009), one EDS course at the JVI (2007), and several regional seminars, namely on ITRS (Baku, 2007), CDIS (Ljubljana, 2008), International Trade in Services (Kyiv, 2008), and Remittances (Ljubljana, 2009).

61. The participants in general found the courses beneficial to their work. They benefited from the workshops and discussions on methodology during their training. Concerns were raised over the limited opportunities for the MFS training going forward, which are further constrained by language. In this connection, the feasibility of the NBG hosting a Russian-language FSI regional seminar was discussed.

TA needs for the period ahead

62. Despite the achievements in MFS, TA is needed for Georgia to begin to compile financial soundness indicators. There is also need for traditional training in MFS, in compilation issues based on the Compilation Guide, 2008, and in specialized training on FSIs (preferably in Russian).

63. While STA’s TA has helped the NBG make substantial progress in the area of external sector statistics, further TA is envisaged for strengthening data sources, notably the ITRS and private external debt compilation program. Training is requested by the authorities in methodology of the Balance of Payments and International Investment Position Manual (BPM6), and in compilation issues for the international trade in services statistics.

C. The National Statistics Office of Georgia

64. GEOSTAT was established as an independent agency in January 2010, as the successor to the Statistics Department of the Ministry of Economy of Georgia. The new management of the institution is giving priority to building trust and increasing the credibility of the statistics produced by GEOSTAT. Much remains to be done.

65. A number of key lessons can be derived from the experience of GEOSTAT and its predecessor institution in absorbing TA. These include: (1) the mainstreaming of statistics in the national priorities, and linking them to the budget; (2) the need for extensive coordination with donors; (3) the importance of a statistical legal framework; and (4)

the central role of adequate resources, including having relatively limited staff turnover and the means to facilitate the transfer and retention of knowledge.

Modalities of TA delivery and major outcome

66. During the period 2005 to mid-2010, STA delivered one TA mission in national accounts (2005) and three missions in price statistics (2005, 2006, and 2007). In general, the effectiveness of the TA delivered to GEOSTAT was hampered by the lack of financial, staff, and computer resources, as well as staff turnover and unresolved communication issues within the institution.

Real sector statistics

TA delivery and outcome

67. The July 2002 Data ROSC assigned seven largely nonobserved (LNO) ratings to the national accounts statistics, a highly critical evaluation, including on the legal basis and resources for statistics, source data and statistics techniques (inadequate coverage of the business register; poor coverage of units in terms of value added for agriculture, retail trade, construction, catering, and services; limited administrative data sources to estimate the nonobserved economy; and inadequate data for imports and exports of services), and data dissemination practices.

68. As a result of the IMF’s TA program, and technical and financial assistance provided by other TA providers, quarterly and annual GDP estimates are produced according to the System of National Accounts 1993 (1993 SNA) under both the production and expenditure approaches. The national accounts and price statistics are reported by GEOSTAT to STA using the ICS on a timely and regular basis for publication in IFS. The 2004 and 2005 TA missions that followed the ROSC found that, although the authorities had made good progress in addressing some of the issues, there was scope for improving the constant price estimates and advancing the development of a system of SUTs.

69. The TA output in NA statistics has been affected by the relatively weak legal framework in place, and by inadequate GEOSTAT resources. The reporting of statistics is voluntary; no sanctions for nonreporting or misreporting are stipulated by the Statistics Law. In 2009, the GDP data were revised, based on the results of a business survey. However, there is a need to update the sample frame for the quarterly agriculture sample survey, which is based on the 2004 Census of Agriculture, although progress is mostly dependent on donor financing. There is also a need to improve basic statistics. The lack of IT resources hampers the development of business statistics, which are based on the quarterly census of big companies and a sample of small and medium enterprises. Overall, the compilation methods for the volume measures have remained largely unchanged. There is no GDP by expenditure approach at constant prices due to the absence of the export/import price indexes and problems for deflation of consumption. Consumption of fixed capital has now been estimated with the historical time series back to 1996. A set of SUTs has now been prepared in more detail than the one compiled annually for the years since 1998. Household consumption, which was published earlier as a total figure, is now broken down according to the COICOP classification.

70. On price statistics, progress has been achieved regarding the development of agricultural price indices, and there are plans to upgrade both the CPI and PPI methodologies. The PPI specification and dissemination practices, in particular, still need improvement. The 2007 mission focused on the progress made on extending the scope of the index to cover the agriculture sector and emphasized the importance of conducting the full survey, at that time expected to be financed by TACIS. The collection sources and methods were evaluated, and a detailed analysis was conducted of the price data obtained from the pilot survey over the July-December 2006 period. In addition, the collection methodology, the index structure and weights, and the index series compiled from the pilot survey were scrutinized. The treatment of seasonal products in the CPI has been strongly criticized by users, and remains to be addressed. A new procedure, involving the imputation of out-of-season prices, was implemented in January 2007 as part of the index rebase, however there is still room for improvement. No progress so far has been reported on the PPI, which still has a number of gaps, including roads and the preparation of constructions sites.

Training for GEOSTAT staff

71. Extensive training has been offered to GEOSTAT staff by STA. GEOSTAT staff attended 11 various training activities in real and external sectors statistics including NA training courses at the INS (2006 and 2009), quarterly NA course at the JVI (two staff, in 2008), and price statistics course at the JVI in 2006. In addition, as a compiling agency for several components of the external sector statistic, GEOSTAT staff was accepted to the BOP statistics courses (INS in 2005, JVI in 2006 and 2008), external debt statistics course at the INS (2007), and regional seminar on remittances held in Ljubljana in 2009. The participants noted that the longer NA courses on methodology were more beneficial (since they laid a better foundation) than the shorter ones.

TA needs for the period ahead

72. The TA priorities are to be defined in the three-year Master Plan, which is currently under development. TA in quarterly NA volume measures, and further development of SUTs is needed. The IMF’s TA program for GEOSTAT should be coordinated with other TA providers (currently the World Bank and the European Commission). Within the framework of their restructuring and reshaping, GEOSTAT management expressed the need for more training in NA and price statistics.

D. The Ministry of Finance of Georgia

73. The MoF is responsible for compiling government finance statistics. Data are reported to STA for publication in the GFS Yearbook and in IFS. Consistent time series are available for three years due to transition to GFSM 2001 methodology. There is clear understanding of the synergies between the quality of GFS statistics and implementation of the accounting reform, strategy of which is set out in Decree 101 issued by the MoF on February 10, 2006. Reforms are documented in the paper Accounting Reform Strategy 2007-2017.

74. The 2002 Data ROSC found that fiscal statistics in Georgia at that time did not meet international standards for concepts, definitions, and classifications. Although the analytical framework for fiscal statistics was broadly based on the GFSM 1986 methodology, there were important departures from that methodology. The largely not observed ratings were also assigned for resources, transparency, and relevance of the GFS.

Modalities of TA delivery and major outcome

75. The GFS TA mission in 2006 noted considerable improvements in GFS data. By

the time of the TA mission, Georgia had embarked on a program of reform to its central and local government budget and accounting systems with a view to fully adopting the GFSM 2001 methodology and the International Public Sector Accounting Standards (IPSASs). Nevertheless, the degree of implementation of STA’s recommendations in this area have been somewhat constrained by the requirements of the budget laws and the pace of implementation of the accounting reform.

76. The Treasury single account has been established, cash planning and cash management introduced, and GFSM 2001 functional and economic classifications are being applied to the modified-accrual data. Specifically, the basing of the classification of outlays on the Classification of Outlays by Functions of Government (COFOG) (GFSM 2001 version) was introduced in the January 2008 for Central Government and from 2010 for Local Governments. Unresolved issues remain with regard to the classification by institutional sector of the entities set up by government as Legal Entities at Public Law (LEPLs), the valuation and coverage of nonfinancial assets, and the need to expand balance-sheet items (currently, only liabilities are compiled, which are classified in accordance with GFSM 2001). There have been noticeable improvements in the accessibility and serviceability of GFS, with expanded information on the upgraded MoF website. The MoF was generally satisfied by the modes and outcome of the GFS TA, and expressed a preference for assistance through a peripatetic advisor.

Training for MoF staff

77. During 2005-2009, six staff from the MoF attended GFS statistics courses at the JVI and INS, and one person attended an External Debt course at the JVI. MoF officials considered STA’s training activities in GFS very effective.

TA needs for the period ahead

78. The MoF expressed a strong desire to increase Georgia’s access to GFS training.

The MoF feels that further TA needs will be determined at a later stage of its reform program, subject to the pace of reform implementation.

V. Conclusions

A. Albania

79. The substantial recent increase in staff resources for INSTAT, particularly at a time of budget cutting, is an important demonstration of the commitment of the Albanian government to the development of macroeconomic statistics. 47 new staff are being allocated to INSTAT—30 are compilers and nine are to work on national accounts. Together with recent organizational changes that have permitted salary increases for staff, this should improve the quantity and quality of staff working on national accounts, thereby addressing a weakness that undermines progress in enhancing statistical capacity in many countries.

80. Albania has made substantial progress in statistical development in recent years, including especially in areas covered by STA’s TA, but the protracted delays in its subscription to the SDDS make it increasingly an outlier. Albania had been targeted as a potential candidate for early SDDS subscription for much of the previous decade, but road blocks remain and seem unlikely to be resolved at a very early date.

81. TA to Albania seems to have been broadly successful. The Data ROSC of 2006 provided a mixed picture of the state of statistics in Albania, with the biggest problems on the national accounts side. National accounts were therefore the main focus of STA’s TA efforts, with a peripatetic advisor making repeated visits over a two-year period. INSTAT officials were very positive about his work, which covered the full range from design of strategic priorities to hands-on assistance with actual compilation. The expert also was involved in advising INSTAT on priority areas for the follow-on assistance being provided through the EU (see below). A wide range of statistical improvements can be identified with his work.

82. Progress has been achieved also across a number of the other topical areas, serviced through conventional short-term (two-week) missions. Most particularly, a number of surveys on external statistics identified as needed in the Data ROSC have been conducted. On the other hand, there has been no progress on some issues identified as priorities in the Data ROSC, i.e., enforcing penalties for nonresponses to surveys.

83. Insofar as there have been shortfalls, this has been ascribed to a lack of an overall coordinating and prioritization framework. As in many countries, coordination between the various producing agencies represents a continuing challenge. Full buy-in by senior officials of the various agencies could enable a significant acceleration in statistical development.

84. There was universal appreciation of the value of IMF training in statistics. The IMF remains the sole source of training in many of the macroeconomic statistics areas, and IMF training is seen almost as a prerequisite for holding any position beyond a relatively junior level. Particularly at a time of rapid build-up of resources, such as INSTAT is experiencing at the moment, lack of rapid access to training may cause a road block in statistical development.

85. There remain also important issues of lack of credibility among at least some of the compiling agencies. At the users’ meetings, and confirmed at meetings with officials, there were strong indications that Albania’s history of past government involvement in statistics has generated a legacy that is not yet fully resolved that there is mistrust in official statistics. As in other countries, issues such as whether the country has gone into recession have threatened to produce an apparent politicization of statistics, meaning that whatever numbers are produced may not be believed. While Albania has introduced an institutional framework that seeks to mitigate these concerns—the BoA for instance has statutory independence and INSTAT was taken out from its former ministry and placed directly under the Prime Minister, the establishment of credibility seems not yet complete, and might be an area for TA. In particular, credibility might be helped by further increases in the transparency of statistics, such as disseminating fully the weights in the baskets for the CPI, quarterly GDP in current prices, and other indicators: the reasons for any revisions in any statistics; and meticulous adherence to ARCs for statistics.

86. In the period ahead, EU-sponsored programs will be providing the largest part of TA in statistics to Albania. Albanian agencies have been twinned with a number of EU counterpart agencies, and extensive programs prepared for implementation over the coming few years. This raises questions as to what role (if any) remains for STA. The various agencies, for their part have all expressed keenness to maintain an active STA presence. In many areas training is seen as the priority, with a possible division of the work so that the EU agencies provide the TA and STA the training. Even with this broad division, however, there are a number of areas where continuing TA is likely to be sought from STA: first, insofar as STA TA is linked to issues of Fund surveillance; second, on issues related to “quality” and other aspects of the IMF’s standards work, including subscription to the SDDS; and thirdly, areas where STA’s worldwide expertise, in particular on issues related to statistics in transition and developing countries, might fill gaps that are not covered from the experiences of the central banks and statistics agencies in long-developed countries that are twinned with the Albanian agencies. Overall these residual needs may be significant. The mission and the officials of all the statistics compiling agencies agreed that any such assistance should be coordinated with that provided through the EU.

B. Georgia

87. There have been significant changes in recent years in the institutional arrangements governing the production and dissemination of statistics in Georgia. Most significantly, at the beginning of 2010 the Department of Statistics, formerly under the auspices of the Ministry of Economic Development, was re-established as GEOSTAT, an

autonomous agency responsible directly to the President. In its new role there is considerable operational independence, and in principle GEOSTAT is no longer bound to pay only civil service wage rates. GEOSTAT is however still subject to the budget for its financial resources. These have been cut by a cumulative 50 percent in recent years, leading to wage rates that are below those for the public sector as a whole, as well as difficulties in staff recruitment and retention: GEOSTAT has been mandated to make up its resource cuts through generating revenue through charging for its products. At the same time, there is reportedly still a negative public perception of GEOSTAT, apparently a legacy of its earlier structure and former management. Absence of credibility in official statistics may serve to impede the effective implementation of policy.

88. STA technical assistance seems to have been broadly successful, judging it against the baseline of the results of the Data ROSC conducted in 2002. At that time, requirements in most categories in all topical areas were only “largely observed” or more seriously “largely not observed (LNO).” Georgia was not even participating in the GDDS and, unusually for any Data ROSC, there was not even mention of the SDDS even as a medium-term goal. Area department views damningly were that balance of payments data were not even broadly adequate for surveillance. In recent years, in contrast, Georgia has adopted the ITRS, subscribed to the SDDS, and introduced multiple other statistical reforms. Many of these can be directly related to the TA delivered by STA.

89. Technical assistance was most intensive in external sector statistics, where for two years a regional advisor was based in Azerbaijan working exclusively on Azerbaijan, Georgia, and Uzbekistan. Additionally, there were since 2005 STA missions on national accounts, price statistics, as well as monetary and government finance statistics. Where clear success results were observed these could be related to:

  • The availability of committed counterparties;

  • Low staff turnover;

  • Clearly specified target objectives.

90. An important reason where TA was not successful seems to have been lack of resources, in particular that TA was delivered on the presumption that donor financing was essentially in place. When this financing fell away, the agency concerned—GEOSTAT—was unable to take implementation forward. There was also some limited criticism that TA had occasionally not been sufficiently hands-on to be operationally useful.

91. In addition to assisting with the implementation of technical enhancements in Georgia’s statistics, STA also assisted Georgia through its coordinating role; in particular, as regards the SDDS, Georgia’s progress, like that of several other potential SDDS subscribers, was stalled over issues of coordinating and monitoring the country’s commitments upon SDDS subscription. In the event, an STA TA mission helped craft an arrangement between the NBG and GEOSTAT under which the latter would retain the role of SDDS coordinator, in line with its overall responsibilities for statistics in Georgia, while the NBG would carry out the resource-intensive activities of maintaining the ARC and monitoring compliance with the ARC.

92. There was a general feeling among the Georgia agencies that the modes of delivery had been broadly appropriate. The intensity of the assistance from the regional advisor was seen as well suited to the need to address the pervasive problems of the BOP, while one-off short-term missions had been well suited to address specific issues in the other datasets, i.e., broadening the institutional coverage in the monetary statistics. Looking forward, it was generally agreed that, for the time being at least, short-term single-focused missions would likely be sufficient to meet Georgia’s needs.

93. There was some concern that Georgia might to some extent fall between the cracks in its access to TA delivery, in that, while the Fund has substantially expended the role of its Regional Technical Assistance Centers (RTACs) in delivering TA, Georgia and some of its near-neighbors are at present not designated to participate in an RTAC. Beyond this, with focus on the “new” statistics identified in G-20 work and as a result of the global financial crisis, less prioritization might be given to a small country such as Georgia at the incipient stage of becoming an emerging market economy. This would be particularly problematic since in some of the areas in which the Fund has hitherto provided assistance, in particular monetary, balance of payments, and government finance statistics, there are no comparable suppliers.

94. There was universal agreement as to the importance of IMF training in statistics. Many of the key counterparts to the evaluation mission had themselves been on such training, and stressed how much it had benefitted them. It was widely felt that Georgia’s training needs were substantially greater than what was presently available. Georgia has access to only the JVI, among the regional training centers, and the highly selective eligibility for INS courses in Washington. Both the NBG and GEOSTAT expressed serious interest in hosting a regional training course on a specific statistical area in Tbilisi next year.

95. Among the key findings of the mission were:

  • STA’s TA was well designed and broadly successful; where it had not been successful, this seemed due to the serious lack of available resources.

  • Going forward, there are, in each of the three agencies, a number of technical issues in prospect where further targeted STA TA would be useful. The MoF, for instance, has encountered a number of glitches in implementing its plan to migrate to GFSM 2001 by 2017, and TA might be useful to ensure that this plan stays on track. More generally, now that progress has been made and Georgia has subscribed to the SDDS, further statistical enhancements could buttress Georgia’s position at an important stage of becoming an emerging market economy.

  • Where the successful implementation of TA is likely to require additional resources, particularly as regards GEOSTAT, STA could work with the agency to identify likely donors and to jointly design projects that would include the donors’ participation.

  • Enhancements to the availability of training should be explored, perhaps through assisting the NBG or GEOSTAT to host regional training.

  • While SDDS subscription should serve to substantially enhance the credibility of Georgia’s statistics, further efforts may be useful. For instance, Georgia’s Data ROSC is now very out of date and conveys a negative impression as to the quality of Georgia’s statistics: an update to the Data ROSC may be useful. It may in this regard also be useful to consider direct assistance as to GEOSTAT’s communications strategy.

  • The recent increase in staff resources at INSTAT, particularly coming at a time of government cutbacks, is a strong sign of official support for macroeconomic statistics at this time.

C. Overall

96. The evaluation indicates strongly that STA’s TA was successful in working on the priority issues and using appropriate modalities to address them, in particular peripatetic/regional advisors for difficult topics of major concern. Albania and Georgia are both transition economies, and have charted different courses to their present state of statistical development. STA did not apply one size fits all, but provided different forms and coverage of TA depending on the two countries’ circumstances.

97. The Data ROSC exercise was useful, among other reasons, for identifying the key issues to be addressed, both through the ratings of key attributes of data quality, and through the prioritized recommendations included in the Data ROSC. Both Georgia and Albania have demonstrated substantial implementation since the Data ROSCs conducted in 2002 and 2006, respectively. For Georgia, in particular, the Data ROSC is now very much out of date, and would be misleading if used as a guide as to the present state of Georgian statistics. It would be well worthwhile to update this Data ROSC, in part to provide an accurate contemporary picture of Georgia’s statistics, but also to give additional guidance as to TA needs for the period ahead.

98. Peripatetic/regional advisors look to be a useful means to supplying technical assistance, at least to countries in a similar state of statistical development as were Albania and Georgia. A substantial work program can be covered over a period of a year or two, the agencies and the experts get to know each other well, and develop an understanding as to how best to work together, and the periodic absences of the experts encourage knowledge transfer and the operationalizing of the knowledge by the agency concerned.

99. Training is a critical part of enhancing statistical capacity, and of sustaining that enhancement. There seem to be few alternatives in most areas of macroeconomic statistics to training by the Fund. In both countries, demand seems to be well outstripping supply. This calls for considering new approaches to complement the traditional approach to training: perhaps more in-country courses (Georgia showed serious interest in hosting a regional training course); emphasis on recipients of training passing on the training (train the trainers); or possibly some distance learning.

100. A medium-term focus for the TA would be useful. For Albania one focus might be on SDDS subscription, so that road blocks could be identified and a critical path analysis be conducted so that the road blocks be bypassed and the country can achieve subscription, and other important milestones, as soon as possible. Such an approach might also facilitate interagency coordination where there are many suppliers (see below).

101. TA might usefully be offered not only on the macroeconomic topics, but on the other aspects of data quality, especially since these often determine whether a development strategy succeeds or fails. Among the areas where STA’s expertise could be useful are legal and institutional arrangements, and enhancements in transparency and work practices (e.g., on revisions policies and advance release calendars).

102. In some countries (such as Georgia), STA may be almost the only supplier of TA in statistics; in others (such as Albania) it may be at least for a while a relatively small player. Even in the latter, there may well be significant areas where STA’s TA is to be provided. This leaves a policy decision: should STA run down its assistance relatively in a country where there are other providers, or should it use the opportunity of the presence of the other providers to focus the country’s share of TA more intensively on areas remaining under STA’s auspices. Probably the STA strategy for such a country will end up as a compromise between the two options, but continuing to look at each country on a case-by-case basis, and coordinating with the other providers, are likely to be important elements for moving forward.