Of the Statistical System of Kazakhstan

1

A Global Assessment

of the Statistical System of Kazakhstan

Final version, 27 February 2004

CONTENTSPage

Preface 5

Executive Summary 6

A. Institutional Setting 6

B. Statistical Capacity in Major Fields 8

A. Institutional Setting12

1. Legal Setting12

1.1 Statistical Law12

1.2 Other Legislation13

1.3 Statistical Councils, Statistical Programmes, Inter-ministerial 15

Consultations, Consultations with Users

1.4Protection of Independence17

1.5 Data Protection and Confidentiality18

2. National Statistical System19

2.1 Principles19

2.2 Structure of the System20

2.3 Liaison with Research Institutes and Universities25

2.4 Liaison with the Business Community25

2.5 Product and Dissemination Policies26

2.6 Relations with the Central Government28

2.7 Relations with Local Governments29

2.8 Access to Administrative Records30

3. Statistical Agency of the Republic of Kazakhstan31

3.1 Mission, Internal Organisation31

3.2 Planning, Programming and Priority Setting33

3.3 Monitoring of Quality in All Activities34

3.4 Finance and Budgeting36

3.5 Staffing37

3.6 Information Technology39

3.7 Internal Monitoring of Performance41

3.8 External Accountability 41

B. Statistical Capacity in Major Fields43

4. Overall Capacity43

5. Classifications43

6. Registers45

6.1 Business Register45

6.2 Population Register47

6.3 Territorial Units Register48

7. National Accounts48

8. External Trade Statistics50

9. Prices52

9.1 Consumer Price Index, CPI52

9.2 Other Indexes53

10. Population Census54

11. Household Surveys54

11.1 Labour Force Survey54

11.2 Income and Expenditure Survey55

12. Annual Population Statistics56

13. Business Surveys and Related Short-term and Structural Statistics56

14. Service Statistics58

15. Surveys of Agricultural Holdings and Related Agricultural Surveys59

16. Public Sector Activities60

17. Monetary, Financial, and Balance of Payments Statistics60

17.1 Monetary and Financial Statistics60

17.2 Balance of Payments Statistics61

C.Recommendations62

18. General recommendations in organisational and subject matter areas62

19. Recommendations regarding the prospective fields of co-operation

within future Tacis programmes65

Annexes67

1. Law on State Statistics 67

2. Organisation of AoS, incl. regional structure 74

3. Programme to Improve Official Statistics in KR for 1999-2005 76

4. Staff, number, age structure, education 113

5. Schedule of press releases 115

6. List of main publications 116

7. Meetings held during assessment missions 118

8. List of persons met 124

9. Summary of user survey results 126

Preface

The Global Assessment of the Statistical System of Kazakhstan was carried out by the following two experts: Ms Marie BOHATA, Director Institute for Governance and Leadership in the CzechRepublic, and Mr. Gösta GUTELAND, consultant, Sweden. The assessment was focused on describing and analysing, from a horizontal point of view, the present situation in the Agency of Statistics of Kazakhstan (the AoS), and other institutions in Kazakhstan that are responsible for production of statistics.

The main objectives of the global assessment have been the following:

To assess the administrative capacity of the country in the field of statistics,

To assess the technical aspects of the statistical system in the country,

To assess the compliance level of Kazakh statistics with European statistical requirements, pointing out the influence of the Tacis assistance in this regard,

To prepare recommendations regarding the prospective fields of co-operation within future Tacis programmes

During the first mission (12 – 23 May 2003), visits were made to the AoS, the National Bank, the Academy of Economics and Statistics in Almaty and two regional offices. There were also meetings organised with representatives from the EC-delegation. During the second mission (7 – 11 July 2003), a draft report was discussed with the AoS. Visits were also made to the Ministry of Finance of Kazakhstan, the Ministry of Trade and Industry and the Customs Control Agency in Astana.

The assessors were very well received and helped by the AoS and the visited institutions in the collection of facts about and plans for the development of the statistical system in Kazakhstan. Discussions were held in an open and friendly atmosphere. Without these contributions and without the excellent translations that were provided it would not have been possible to carry out the mandate in the short period that was given for the assessment.

Executive Summary

A. Institutional Setting

The Law on State Statistics provides the legal basis of the national statistical system of the Republic of Kazakhstan (RK) and sets out the terms of reference of the Agency of Statistics (AoS) at its core.

The current Law on State Statistics was adopted by the Parliament in 1997 and amended in January 2002. Based on the UN Fundamental Principles for Official Statistics, it states the right of the AoS to conduct statistical surveys, to elaborate programmes for statistical observations, to collect statistical data from natural and legal entities, to check the reliability of data and to verify the primary information. The AoS is authorised to request the information it deems necessary from all public and private institutions, as well as from individuals. All respondents must furnish the requested information accurately, and in the format and time period determined by the Plan for statistics.

The AoS is responsible for statistical classifications and nomenclatures, and using the powers provided by the Law it co-ordinates the whole statistical programme of the country. In fulfilling the co-ordinating role, it is assisted bythe Inter-Agency Council for Improving Government Statistics, which is chaired by Deputy Prime Minister. The AoS also has the duty of methodologically approving all surveys conducted by Ministries and other bodies involved in the state statistical service. It has formal agreements on mutual collaboration and data exchange with most of the institutions in the state statistical service and based on the Law on State Statistics, has unrestricted access to administrative sources of information.However, certain restrictions might be found in other Laws.

The AoS is an independent government institution in charge of the state statistics policy. Law on State Statistics ensures AoS independence in the field of methodologies and statistical production. The Chairman of the Agency and his Deputies fill their political civil posts. Staff members of the AoS hold administrative civil posts.

The mission of the service embodies the commitment to meet statistical requirements of the government as well as of the citizens.

The Law on State Statistics stipulates such principles as integrity, reliability, and sufficiency of statistical indicators, comprehensive and unbiased study of economic and social development, accessibility and openness, as well as comparability of applied methodology.

The Republic of Kazakhstan has not adopted a Law on the Protection of Personal Data. If this happens, important implications for the work of the AoS, as well as for that of almost all other bodies in the RK may be expected. Confidentiality of primary statistical data is guaranteed for legal and natural persons by the Article 13 of the Law “On State statistics”.

There are various universities and research institutes in the RK closely co-operating with the AoS. There is a long-term, wide-range and very productive co-operation between AoS and the Academy of Economics and Statistics of the city of Almaty (AESA). AESA being the scientific institute of the AoS, is responsible for the vocational training of the AoS staff both in the field of statistics and in the field of information technology. It also supports the statistical departments of the AoS in developing new statistical methods.

Since the creation of the independent Republic of Kazakhstan, multi-annual programmes serve as a basis for the development of state statistics. From the long-term programme,annual planson statistics improvement are derived; these plans are approved by the Chairman of the AoS. Besides, annual AoS plans covering all statistical surveys conducted by the AoS are elaborated.

Main users are the Administration of the President, the Parliament, Government, National Bank, ministries and agencies and local authorities. Gradually, the interest of other user groups covering enterprises, financial organisations, scientific and education institutions, libraries, international organisations, mass media, non-governmental institutions and the general public has been growing.

Since 1997, Tacis assistance in dissemination and publication of statistical information has allowed for remarkable improvements in publication policy and deliveries. The publication programme was expanded, the structure, content and design of publications improved, and compliance with international standards achieved.

The structure of the national statistical system as well as the organisational structure of the AoS reflects the complexity of the institutional administration of the country. Since 2002, the administration consists of the central (Republican) government, 14 oblasts (provinces) plus the cities of Astana and Almaty that have oblast status; oblasts are divided into 160 rayons (districts). There are 39 towns with oblast and 45 towns with rayon status,. The AoS consists of 8 departments, 16 subordinated regional offices (RSO) having the status of State bodies and 16 subordinated regional centres of statistical information (CSI) that do not possess the status of State bodies. Subordinated institutions at the national level are the following: the national state owned enterprise ”Information and Computing Centre ” responsible for collecting, checking, capturing and processing of statistical data, as well as for the information technology of the AoS, the joint-stock company ”Academy of Economics and Statistics of the City of Almaty“ ( recently privatised) responsible for the vocational training of the AoS staff in the fields of statistics and information technology, and development of new statistical methods, and the formerly State owned enterprise ”Kazstatinform” (recently privatised).

The structure of the statistical system at the regional level (oblasts) comprises three entities involved in the whole statistical process: the Regional Statistical Offices (Government offices), which are in charge of dissemination and analysis of information at the regional level, Centres of Statistical Information (public institutions that are not Government offices), which are in charge of data collection and processing, and the Information and Computer Centre establishments (branches of the RSE “ICC of the Agency of Statistics of RK”) , which develop information technologies at the regional level and are authorised to disseminate information on the basis of payment.

Since the existing organisation is complicated and some responsibilities are not sufficiently clear, a new organisational structure has been designed. Main features of the new organisational structure are: liquidation of Information and Statistical Centres at regional level and re-allocation of employees according the current regional needs and an increase in the number of statisticians at the Agency of Statistics of RK. Thus, the new organisation would encompass two parallel hierarchies of the state agency (Agency of Statistics of RK) and State enterprise (Information and Computer Centre of the AoS), both composed of three levels: the centre, the region and the rayon.

The number of personnel of the AoS organisation is limited to a maximum of 3955 staff members ( plus 2319 rural statisticians working at rayon –district- level in rural areas but directly subordinated to the regional –oblast- statistical offices), out of which about 86% work in the regional institutions. 90% of all staff are women; in the central office of the AoS they represent about 84%. The average age of the staff members is about 41 years. About two thirds of all staff members are graduates (institutions of higher education, universities), this ratio is higher in the central office than in the regional offices.

IT equipment is modern and operated by well-skilled staff. However, this will not continue automatically, and thus, a plan of renewal of IT equipment should be established.

A lot of progress has been achieved in Kazakh statistical service and quality of statistics has improved considerably. At this stage, the AoS sees some urgent tasks, which are conditional for further quality enhancement: improvement of the organisational structure, development of new statistical fields, strengthening of the scientific character of statistical activities and usage of administrative sources of data. In this respect, a need to set up a scientific institute focusing on statistical research and forecasting is being considered.

The design of the system aiming at production of a very high number of detailed indicators and regional, often also district coverage of the country and high frequency is very ambitious. It assumes extremely good discipline of all respondents whose burden is significantly bigger than in other countries. The overwhelming majority of surveys are exhaustive, full coverage surveys. Adoption of sampling techniques in order to reduce costs and the burden put on respondents is highly recommended. For this purpose, the basic statistical tools have been already developed (mainly the business register), but the main problem still remains: the move from a functional to an institutional approach. This move means a change in the prevailing philosophy of the system based on describing branches (economic activities) rather than economic actors.

Internal monitoring of performance, as well as external accountability of management and staff of statistical institutions are still to be achieved. A systematic monitoring of quality should be implemented. This system should cover the actual scope and coverage of statistics as well as consistency and timeliness.

B. Statistical Capacity in Major Fields

Tacis technical assistance to Kazakhstan started in December 1992. At that time the statistics in the country was restricted to collection of data to be delivered to Moscow. Substantial progress has been made during the different phases of co-operation that have so far been carried out.

In general, the statistical capacity in Kazakhstan can be regarded as good. Although there are no systematic monitoring the quality seems in most cases to be satisfying. There are, however, some special strengths and weaknesses.

To the positive side could be counted the high coverage of statistics that are normally included in a national statistical system. A modern system of classifications for statistics has been adopted and to a large extent implemented. High response rate in surveys could also be counted as strength. This is of special value for the Labour Force Survey and for the Household Income and Expenditure Survey. In general economic statistics is of good quality, even if there are some exceptions. The living conditions for the population are comprehensively described. On the positive side the many nice publications could also be counted for. A well developed Website is used for dissemination. There is also a brand new library, which will be of great value for many users. The Tacis projects seem to have given a substantial contribution to the progress and the standard of the national statistical system. The projects seem to have had a special positive impact on the dissemination system and on the development of sampling.

On the negative side a low capacity at the national level might be noticed. The organisation is now very much dependent on a few key persons at each subject matter field.

For specific topics some of the weaknesses could be summarised in the following way:

1. The business register ought to be cleaned from inactive enterprises. The yearly questionnaires from the local statistical offices and the register at the Tax offices ought to be used for this work. The need for a costly Business census is reduced if the existing sources are used in an efficient way. The use of other sources for the updating of the Business register will probably be much more successful if all concerned authorities are using one unified code for the classification of enterprises.Due to the fact that the standard statistical units in the Business register are not always used for sampling the results from surveys based on sampling are not as reliable as they ought to be. Questionnaires are designed for branches instead of for complete statistical units.

2. There is an urgent need for a continuously updated population register in Kazakhstan. Priority ought to be given to the work with the register on population at the Ministry of Transport and Communication. The work with the register has been delayed due to uncertainties about who should be the principle customer. That problem ought to be solved as soon as possible. This register should be used for administrative purposes. At the AoS a statistical population register ought to be developed. In order to avoid duplication the work with the two registers should be closely co-ordinated. The statistical population register should only be used for statistical purposes.

3. National accounts are in principle produced and published according to international standards. There seem however to be a risk for an underestimation of GDP as the non-observed economy still is very large.

4. Although substantial improvements of the quality of external trade statistics have been achieved, export and import volumes continue however to be underreported by unknown amounts. There should be a more clear division of labour and a closer co-operation between the AoS, the Customs Agency and the National Bank in order to achieve improvements.

5. The CPI is calculated according to international standards. The ambition is very high with both weekly measurements and a very detailed division of data (consumer prices are registered in each rayon centre). For the future it might be possible to abolish the weekly measurements. They were once introduced when the inflation was very high.

6. The Balance of Payments seems to meet all main international requirements in the field. The IMF Manual of Balance of Payments’ Statistics gives the methodological framework. The most urgent need for improvements is the estimates of the non-observed economy. There is also a need to improve information on State Finances.

7. The Census 1999 was carried trough in a successful manner. The AoS had the capacity for such a work. The lack of a proper population register might make it necessary to use the census from 1999 for sampling for a long period. In the worst case it will last as long as until the next census in 2009.

8. The Labour Force Survey follows international standard. The ambition is very high. To produce and present quarterly figures at oblast level is very demanding on resources in the long run. It is important to receive resources for maintaining the standard. To some degree the quality of the LFS is depending on the quality of the population register from which the samples will be drawn.

9. The sampling frame for the household income and expenditure survey (HIES) is also based on the Census 1999. The sample size for the HIES has been steadily increasing. In 2003, the household statistics accounted for 19% of the expenses of the regional Information and Statistical Centres. An appropriate sampling technique is used to select the number of families representing each income level from each region of the country. Geographic coverage is also comprehensive as the survey includes representative samples of urban and rural households. As for the LFS there might be a problem with the sampling when the census data no longer reflect the actual population.