Republic of Tajikistan: Detailed Assessments Using the Data Quality Assessment Framework (DQAF July 2003)

This Report of the Observance of Standards and Codes (ROSC) Data Module provides an assessment of Tajikistan’s macroeconomic statistics against the recommendations of the General Data Dissemination System (GDDS), complemented by an assessment of data quality based on the IMF’s Data Quality Assessment Framework. The assessment reveals that Tajikistan already meets many of the GDDS recommendations for dissemination of macroeconomic statistics. The main exceptions are the coverage of national accounts, central government operations, and central government debt. All data categories meet or exceed the recommended periodicity and timeliness.

Abstract

This Report of the Observance of Standards and Codes (ROSC) Data Module provides an assessment of Tajikistan’s macroeconomic statistics against the recommendations of the General Data Dissemination System (GDDS), complemented by an assessment of data quality based on the IMF’s Data Quality Assessment Framework. The assessment reveals that Tajikistan already meets many of the GDDS recommendations for dissemination of macroeconomic statistics. The main exceptions are the coverage of national accounts, central government operations, and central government debt. All data categories meet or exceed the recommended periodicity and timeliness.

I. National Accounts

0. Prerequisites of quality

0.1 Legal and institutional environment

0.1.1 The responsibility for collecting, processing, and disseminating the statistics is clearly specified

The legal basis for the collection, compilation, and dissemination of the statistical information is the Law on State Statistics of the Republic of Tajikistan, adopted in May 1997 and amended in April 2003, to reflect the new status of the State Committee on Statistics (Goskomstat) as an independent state statistics committee. The law includes provisions for all statistical agencies of Tajikistan, but assigns Goskomstat the leading role in the production and dissemination of statistical information. The upgraded status of Goskomstat to a statistical committee, as well as its composition and responsibilities, were spelled out also in the Government Decision of February 2001.

According to the law, an annual State Program of Statistics (SPS) is developed by the Goskomstat in collaboration with the other data-producing state agencies which produce particular sets of statistics. This program includes, among other things, the schedule of surveys to be conducted by the Goskomstat or by other state agencies, the categories of administrative data to be provided to Goskomstat, as well as their schedules of reporting. The agreed draft program is submitted first to the Tajik government, and subsequently to the advisory body—the Statistics Committee of the Commonwealth of Independent States (CIS) for their revision and comments. The final version of the work program is then submitted to the president of Tajikistan for approval. Any additional requests received during the year are centralized and considered for the next year’s working plan.

0.1.2 Data-sharing and coordination among data-producing agencies are adequate

The SPS clearly defines the responsibilities of each state agency for the datasets to be provided to Goskomstat for statistical purposes at the appropriate frequency and timeliness. Data-sharing among statistics-producing agencies and the accuracy of data received is considered satisfactory by the Goskomstat’s managers and staff. The only evidence of duplication of efforts relates to the price statistics, namely the production of a consumer price indicator1 by the National Bank of Tajikistan (NBT), for internal purposes only.

0.1.3 Individual reporters’ data are to be kept confidential and used for statistical purposes only

The Law on State Statistics contains provisions guaranteeing the confidentiality of data provided to the Goskomstat and other authorized state agencies. Article 9 of this Law states that “employees of state statistical bodies must protect the confidentiality of existing statistical information.” However, this provision is ambiguous with regard to the even protection of confidentiality of both physical and legal persons, giving leeway to interpretations. The law might benefit from a clear statement regarding the confidentiality of individual statistical information provided by legal persons.

The Goskomstat enforces confidentiality restrictions. All data relating to information classified as “confidential” (military industry, some of the mining operations, etc.) are transmitted by special courier to the special security section of Goskomstat. This section has the responsibility to (1) protect the confidentiality of certain categories of data and (2) classify data as confidential. Any request for confidential information by department heads or from outside Goskomstat requires the authorization of the security section. For example, in the case of national accounts, special aggregations of data are made in order to protect the confidentiality of such data. Reportedly, there have not been any cases of violations of the confidentiality requirements. Confidentiality of data is appropriately guarded during storage, and the archives are under the responsibilities of a special commission. Staff are subject to disciplinary action for violation of confidentiality.

Respondents are informed in writing of their rights and obligations with regard to the provision of information. At the beginning (cover) of the Goskomstat’s publications, a quotation from the law is included with reference to the use of information for statistical purposes solely.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

The Law on State Statistics provides for mandatory reporting and penalties for noncompliance. The enterprises that fail to report accurate and timely data are liable for fines. In practice, the sanctions are applied only in cases where the enterprises refuse to respond after receiving a warning. The regional offices have the right to impose fines in such cases, however small, but most often the enterprises choose to cooperate rather than to be liable to the board meeting deliberations of the responsible ministry.

0.2 Resources

0.2.1 Staff, facilities, computing resources, and financing are commensurate with statistical programs

Overall, the staff number at Goskomstat is considered appropriate for the current duties, however insufficient in the regional offices.2 Currently, the total staff at Goskomstat consists of 265 employees, of which 166 are working in the computing center. The regional offices comprise 595 staff. Staff has been downsized3 tremendously over the past 10 years (about 70 percent) leading to the reorganization of the institution and restructuring of the work program. Currently, a certain degree of stability has been achieved.

Requests for staff increases have been addressed to the government at times without fruitful results. Owing to the low salary levels, which are not competitive with those of other agencies, it is difficult to retain younger qualified staff. However, there have been recorded cases of staff returning to Goskomstat in recognition of its good work environment.

The national accounts division of Goskomstat consists of five staff, of which three were only recently hired. This division solely is responsible for the compilation of the national accounts. All staff are adequately educated (all are economists). Training of new staff is primarily effected by the head of the division, who explains (1) the general concepts of the methodologies used to compile national accounts and (2) the manner in which the work is organized. Only the head of the division can participate in international courses or seminars on national accounts, provided that the cost is partly or fully sponsored by the organizing institution or other donors (e.g., the head participated in the regional courses in national accounts sponsored by the International Monetary Fund (IMF)).

Computing resources are below needs, particularly in regional centers. For national accounts, there are only two personal computers for the five staff. Furthermore, these computers are equipped with outdated software. However, the processing of raw data (from the different surveys and reports) is done by staff of the computing center of the Goskomstat. About 15 of those staff process data on a permanent basis.

Budget allocation for Goskomstat provided from the government budget is rather tight for the work program. This budget is not sufficient to support development work or computing enhancements. Local offices are supported by the local budget. Financial assistance is sought in order to permit statistical development.

0.2.2 Measures to ensure efficient use of resources are implemented

Management analyzes the work program each year, reviewing priorities. Efficiencies are sought throughout the work process at Goskomstat. Effective collaboration is sought with the other state agencies for the development of the SPS that defines, among other things, the flow of information at the national level to avoid duplication. Reviews of staff performance are conducted with variable frequency depending upon the position (once in two to three years).

Efforts are made to provide staff in regional offices, through periodic training, with clear directions regarding methods and procedures applicable for specific work. Inspections are made regularly (yearly) to check the accuracy of data collection and compilation processes. New possibilities could be explored, nevertheless, for more efficient data transmission mechanisms, use of consistent concepts, and classification across datasets (e.g., use the same classification for household final consumption in national accounts and Household Expenditure Survey (HES), etc.).

In view of the limited resources, tasks are reassigned as needed to even the workload. When one person is absent for various reasons, his or her work is completed by the rest of the team. During peak periods, staff work overtime to ensure the quality of data and to meet the scheduled deadlines.

Expert assistance from international organizations to evaluate statistical methodologies and compilation methods is considered very beneficial.

0.3 Relevance

0.3.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored

Users are kept informed on the current work program, including through the annual catalog of publications, and their feedback is sought. However, the concept of users is limited to a few categories (so-called “principal users”), including government agencies, universities, embassies, and most important business companies. When the principal users receive the catalog of publications, they are also invited to submit requests for publications and types of information of interest and to provide comments as needed in writing. Users’ comments are monitored and responses are provided. Although there is no established schedule for meetings with users, the Goskomstat participates in meetings organized by the government or various ministries, where media and various users are present. Goskomstat is also invited to similar regional meetings organized by the local government units. Goskomstat participates in statistical meetings and seminars organized by international or regional organizations such as the CIS, IMF, International Labor Organization (ILO), and the United Nations (UN).

0.4 Other quality management

0.4.1 Processes are in place to focus on quality

The Goskomstat recognizes that quality is a pillar of sound statistical work, and management is sensitive to all dimensions of quality. The annual SPS is monitored on an ongoing basis, and improvements are sought in all areas. A methodological commission composed mostly of heads of departments (16–23 persons) meets regularly to discuss all proposals for methodological improvements and new developments.

Recent initiatives on quality improvement are the data module ROSC and the country’s decision to participate in the IMF’s General Data Dissemination System (GDDS).

0.4.2 Processes are in place to monitor the quality of the statistical program

A new section dealing with international methods and standards has been set up recently at Goskomstat in view of the country’s participation in GDDS. The quality control of each employee’s work is tightly monitored within each department. Staff are made aware of the importance of quality through internal training programs.

0.4.3 Processes are in place to deal with quality considerations in planning the statistical program

The Goskomstat monitors the quality of its work programs with regard to trade-offs between improvements and available resources. It reviews survey results periodically to identify shortcomings. Feedback from users on newly emerging data requirements is considered, to the extent possible, in planning the statistical program. However, in the case of national accounts and prices, most methodological improvements are the result of the technical assistance provided by international organizations. More internal initiatives for improvement would be very beneficial.

1. Assurances of integrity

1.1 Professionalism

1.1.1 Statistics are produced on an impartial basis

Article 4 of the Law on State Statistics prescribes that the Goskomstat and the territorial statistical services under its control are the official bodies of state statistics and that they gather, monitor, process, analyze, aggregate, disseminate, and store statistical information in accordance with this Law.

Staff is hired on the basis of competence. Vacancies are published in the press and a commission of senior staff of the Goskomstat revises all the application forms to identify the best qualified candidate, after which this commission proposes to the chairman to hire that candidate. Promotions of staff are based on personal performance and proficiency. Moreover, the work environment at Goskomstat supports a culture of competence.

The chairman of the Goskomstat is appointed or dismissed by the president of the republic according to Article 69 of the Constitution, which regulates appointment and dismissal of heads of government bodies.

Staff can participate, to a certain extent, in seminars or courses organized by international organizations.

All publications are reviewed by a commission of senior staff in order to guarantee consistency between statistics.

1.1.2 Choices of sources and statistical techniques, as well as decisions about dissemination, are informed solely by statistical considerations

The Goskomstat is independent in its choice of appropriate source data and statistical techniques. In its publications, the Goskomstat attempts to present the statistics in a way that users understand and appreciate. Statistical data which are considered to be secret are disseminated at an aggregated level with other data categories (e.g., statistics on defense or certain mining operations) although they are not confidential from a statistical point of view. Even though this practice is used to protect confidentiality of individual data, in the above case it is not recommended.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The Goskomstat is entitled to comment and respond on any misused or misinterpreted statistics. When new statistics are published, the Goskomstat explains to the government and other users that the statistics are the result of a process based on international guidelines regarding the collection and compilation rules used.

1.2 Transparency

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The publications and surveys of the Goskomstat contain references and relevant articles of the Law on State Statistics, under which the statistics are collected, processed, and disseminated. Users need previous consent from the Goskomstat to redistribute the statistics contained in its publications. This is mentioned in all publications.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

The government has access to the statistics prior to their release. However, this is not publicly notified since it is “presumed” to be known by the public.

1.2.3 Products of statistical agencies/units are clearly identified as such

All the publications of the Goskomstat are clearly identified by name and logo. When there are joint publications, the national accounts data are clearly identified as products of the Goskomstat.

1.2.4 Advanced notice is given of major changes in methodology, source data, and statistical techniques

Changes in methodology, source data, or statistical techniques are publicly announced in advance in the relevant publications. For example, the change to the methodology of CIS countries to implement the System of National Accounts 1993 (1993 SNA) was announced one year in advance.

1.3 Ethical standards

1.3.1 Guidelines for staff behavior are in place and are well known to the staff

The behavior of staff is governed by the Civil Code. Although there is no formal code of conduct, staff is, upon engagement, informed by management of the rules and guidelines that they have to adopt. Reportedly, top management periodically reminds staff of these rules and guidelines during meetings.

2. Methodological soundness

2.1 Concepts and definitions

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

Goskomstat compiles national accounts based on the methodology developed by the CIS Statistics Committee, which is broadly in accordance with the 1993 SNA. During May and June of 2004, Goskomstat will receive technical assistance, sponsored by the World Bank, to further implement the 1993 SNA.

2.2 Scope

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices

The national accounts statistics consist of the following annual accounts and tables, which are part of the minimum requirements for the implementation of the 1993 SNA, as determined by the Inter-Secretariat Working Group on National Accounts (ISWGNA):

  • Value-added and gross domestic product (GDP) by type of economic activity at current and constant prices (in previous year’s prices);

  • Expenditure of gross domestic product (GDP) at current prices;

  • Annual value-added components at current prices by activity;

  • Primary distribution of income account for the total economy;

  • Secondary distribution of income account for the total economy; and

  • Use of income account for the total economy.

The compilation of volume estimates for the expenditure approach of GDP, as well as the compilation of the capital account, is also a requirement in the ISWGNA guidelines. Goskomstat does not currently compile these data.

The delineation of the constituent units of the economy is generally in accordance with the 1993 SNA. Territorial enclaves in the rest of the world are covered, as well as workers who work part of the year in a foreign country. There are no free zones/bonded warehouses/factories operated by offshore enterprises under customs control, so this is not an issue.

The production boundary is generally in accordance with the 1993 SNA and includes the following:

  • own-account production of all goods for own-final consumption;

  • output of goods for own-account fixed-capital formation;

  • mineral exploration;

  • production of entertainment, literary, or artistic originals; and

  • production of computer software.

Illegal output sold to willing buyers is not in the scope of output measurement.

The assets boundary is broadly in accordance with the 1993 SNA. In particular, agricultural work-in-progress and intangible assets are included in the estimates of gross fixed capital formation (GFCF). Defense-related assets that could be used for civilian purposes and valuables and historical monuments are not in the scope of the assets boundary.

Since 2002, Goskomstat has compiled quarterly accounts (production as well as expenditure approach) experimentally. The results of these estimates, however, are not disseminated.

2.3 Classification/sectorization

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices

Since 1995, Goskomstat has been using the CIS methodology to classify institutional units and transactions. This classification is largely in accordance with the 1993 SNA. However, sectorization of institutional units according to the 1993 SNA was only commenced in 2002 and is currently not entirely completed.

Activities are classified following the former Soviet industry classification, which is not in accordance with the International Standard Industrial Classification of All Economic Activities (ISIC) or the Statistical Classification of Economic Activities in the European Union (NACE). However, Goskomstat is currently preparing the introduction of the General Classification of Types of Economic Activities in Tajikistan, which is largely based on ISIC. Products are classified in accordance with the CIS-recommended classification that was adapted for use in Tajikistan and is neither in accordance with the Central Product Classification (CPC) nor with the Classification of Products by Activity (CPA). The classification used for household consumption is largely based on the Classification of Individual Consumption by Purpose (COICOP), and the classification for government functions is based on the Classification of the Functions of Government (COFOG).

2.4 Basis for recording

2.4.1 Market prices are used to value flows and stocks

The valuation rules used for recording flows and stocks are generally in accordance with the 1993 SNA:

  • Market output is valued at basic prices;

  • Value-added taxes and excise taxes are included in the valuation of intermediate consumption, excluding the deductible part of value-added taxes;

  • Imports are valued on a c.i.f. basis while exports are valued on an f.o.b. basis; and

  • Transactions in foreign currency are converted using the mid-point exchange rate prevailing in the market at the moment they take place.

2.4.2 Recording is done on an accrual basis

Most enterprise transactions are recorded on a cash basis, while all government-related transactions are recorded on a cash basis.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices

Transactions between establishments within the same enterprise are generally recorded on a gross basis.

3. Accuracy and reliability

3.1 Source data

3.1.1 Source data are obtained from comprehensive data collection programs that take into account country-specific conditions

Source data are collected through (1) monthly, quarterly, and annual enterprise reports; (2) annual enterprise balance sheets; and (3) household surveys.

Enterprise reports

Depending on its size, every enterprise has to submit to Goskomstat quarterly or annual reports on its production and intermediate consumption. Data are available for both primary and, if applicable, secondary activity of the enterprise. Each industry has separate reports adapted to its particularities. The most detailed reports (i.e., the annual) consist of four sections: (1) production, (2) intermediate consumption, (3) taxes and subsidies, and (4) inventories. The production section contains, among other things, data on total turnover, value of barter, and own consumption of goods. The intermediate consumption section includes data on the product level, as well as data on consumption of fixed capital. The other two sections list the different taxes paid and subsidies received and the value at the beginning and the end of the period for the inventories of different types of goods (among which, work-in-progress).

All enterprises file monthly or quarterly production reports (depending on the size of the enterprise). These reports contain volume and value data on produced goods both for the current and the previous year. The product classification used is in accordance with a former Soviet classification, although Goskomstat plans to shift to List of Products of the European Communities (PRODCOM) from 2005 onwards.

All enterprises submit monthly or quarterly labor reports. The reports contain data on the number of employees and the total remuneration paid. This concept is in accordance with “compensation of employees” in the 1993 SNA and comprises all payments, in cash and in kind, made to employees (e.g., salaries, bonuses, and premiums in kind payments, and other benefits and social contributions). These reports cover only registered employment while the importance of unregistered employment (and thus wages) is significant. It would be beneficial to conduct a regular focused survey to establish the magnitude of unregistered employment and wages in order to increase the data-based part of estimates of the unobserved economy.

Enterprise balance sheets

All enterprises submit balance sheets to the Goskomstat, which uses them for compilation of GFCF by activity and type of assets. The balance sheets include information on total GFCF, work in progress of construction, software, entertainment originals, and patented entities.

Household surveys

The Goskomstat conducts several household surveys, of which the most important is the HES. The HES is a monthly rotating survey that uses a stratified random sampling technique. The sample frame is derived from census data. All geographic areas and all socioeconomic groups are covered. The sample consists of 1,000 households (of which 625 are in rural areas) that are surveyed by the regional offices. Detailed data on income and expenditure of households are collected through the survey.

The Goskomstat conducts, besides the HES, several other, more focused, household surveys that are mainly used in an effort to quantify parts of the unobserved economy. Examples of these surveys are as follows:

  • Paid services:

    The sample size of this annual survey is about 8,000 households that are randomly chosen by the regional offices of Goskomstat. The survey includes data on purchases of transport, medical, and cultural services.

  • Students:

    This annual survey is conducted with 10 percent of all university students (randomly chosen by the regional offices) and contains data on the payments the students make to their university for registration fees, exams, courses, etc.

Furthermore, Goskomstat collects annual reports from the different governmental bodies to measure their output, intermediate consumption, fixed-capital formation, and final consumption expenditure. Data for some of the government functions (education, health care, social protection, and cultural services) and for the local government are received from the Ministry of Finance (MOF). These data include defense-related expenditures, local government operations, and capital stock information.

Data on imports and exports are received from the Trade Department of Goskomstat and are based on the customs, as well as on direct, declarations from enterprises.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

From a conceptual point of view, the definitions of the collected source data are generally consistent with the definitions, scope, and classification of the national accounts statistics. However, doubts have been raised as to what extent some reported values (of which, among other things, entertainment originals and patented entities) agree with the requirements of the 1993 SNA. It would be beneficial if Goskomstat conducted a study to further examine this issue. Values of consumption of fixed capital are directly reported by the enterprises and are therefore not based on a perpetual inventory method estimate. The data compiled through the enterprise reports and surveys are sufficiently detailed to allow certain adjustments to be made for national accounts purposes. All source data are available on a calendar-year basis in accordance with guidelines for national accounts compilation. Most enterprise data and all government data are recorded on a cash basis.

The coverage of activities (in terms of value added) within the most important industrial groups, including the survey-based estimates of the unobserved economy, is more than 90 percent.

3.1.3 Source data are timely

Data collection and processing timetables are adequate to meet timeliness and periodicity for disseminating the national accounts statistics.

Timely receipt of source data, especially the enterprise reports, is rigorously followed up by the regional offices of Goskomstat. When data are not received on time, Goskomstat contacts the enterprises and urges them to submit the requested data. In case of continued nonresponse, the regional offices of Goskomstat can impose fines.

3.2 Assessment of source data

3.2.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide statistical processes

For all surveys, information on sampling errors and nonresponse rates is not published. Although participating in surveys is on a voluntary basis, reportedly, response rates are close to 100 percent as a result of the fact that respondents do not want to be subjected to administrative follow-up procedures. Therefore, there are no imputations for nonresponse. Individual source data are visually checked for outliers. In case of outlier detection, the submitting unit is contacted to verify the data. The source data are analyzed and checked for temporal consistency.

3.3 Statistical techniques

3.3.1 Data compilation employs sound statistical techniques to deal with data sources

At each stage in the process of aggregation, the data are validated by consistency checks. Grossing-up factors to account for the unobserved economy are derived from several surveys.

3.3.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

The unobserved economy is measured on the basis of (1) several specific sources (regular and ad hoc surveys) that are closely related to the estimated variables and that are regularly conducted, and (2) the number of unregistered employees in an industry. On the basis of survey results, Goskomstat estimates the importance of unobserved activities for several paid services (e.g., transport services and education) and for trade activities by unofficial merchants. For those activities that are not covered by the surveys, an estimate based on unregistered employees is carried out. Therefore, Goskomstat estimates the number of these employees by industry and imputes the average cost of these employees to the corresponding industry’s value added. For the past several years, Goskomstat has estimated the importance of the unobserved economy at between 20 and 25 percent (for 2003, a figure of 24.9 percent was derived). These estimates are incorporated in the estimates of the national accounts. However, it was found that the data used with regard to unregistered employees were incorrectly derived from existing source data. As a consequence, a certain part of the estimates for the unobserved economy is based on faulty data and is therefore unreliable. The exact magnitude of the error will be investigated by Goskomstat. For reasons of consistency, it is recommendable that, once the errors are quantified, current and past GDP estimates will be revised accordingly.

Production approach

Estimates on output and intermediate consumption are compiled at the level of 32 industries. As a consequence of the almost exhaustive coverage of source data, the use of fixed ratios in the estimate of value added is very limited and in no case are the ratios older than five years.

Concerning specific issues related to the compilation of the production approach of GDP, the following observations can be made:

  • The output of the owner-occupied dwellings is valued as the estimated rentals that tenants would pay for similar accommodation. These rental estimates are based on total annual costs connected to owning a dwelling.

  • Data for the work-in-progress are derived from the enterprise reports. They include data on livestock and on large construction projects.

  • In the estimates of output, sales are adjusted for changes in inventories. However, the inventories are not adjusted for holding gains or losses.

  • Data on consumption of fixed capital are derived from the enterprise reports and are therefore not estimated using the perpetual inventory method.

  • Most source data are on a cash basis and are not converted to accrual.

For compiling volume measures of GDP, a single indicator method is used. The enterprise reports contain quantity information on production of the current and the previous year. This data is used to derive volume indices per industry (on the most detailed level, i.e., 32 industries), which are used to compile volume measures of value added. The volume measures of net taxes on products are obtained using the implicit indices of value added of the total economy. A more appropriate procedure would be to extrapolate, at a detailed level, taxes of the previous year by volume measures, that is, by applying the tax rates of the previous year to the volume of transactions subject to taxes in the current year. Volume estimates of trade margins are not compiled. The value added of the trade industry is deflated using the consumer price index (CPI). However, to be in line with international best practices, the volume terms of trade margins should be estimated by extrapolation of the base-year data by volume measures of the sales at a detailed level of the different commodities. Volume estimates are expressed in previous year prices.

Expenditure approach

All components of GDP by expenditure are derived independently. Estimates on household final consumption are compiled on the 3-digit level (based on COICOP), government final consumption is based on COFOG, and GFCF estimates are estimated by activity and type of assets. Data on changes in inventories are derived from the enterprise reports by activity and type of inventories. It is recommended that valuation of inventories be adjusted for holding gains or losses from inventories.

Total estimates of GDP by expenditure do not rely on fixed ratios, derived from benchmarks or other sources, which are more than five years old.

Concerning specific issues related to the compilation of the expenditure approach of GDP, the following can be noted:

  • Government final expenditure excludes incidental sales.

  • Expenses of residents abroad are not included in household final consumption and in imports. Expenses of nonresidents in the economy are not excluded from household final consumption and not included in exports.

  • Expenditure on valuables is not included in capital formation.

Currently, volume measures of GDP by expenditure are not compiled.

3.4 Assessment and validation of intermediate data and statistical outputs

3.4.1 Intermediate results are validated against other information where applicable

Estimates are validated at all stages of processing. Intermediate data on major activities, mainly industry, are assessed against other information, such as volume trends.

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated

Discrepancies in intermediate data are routinely assessed. In case of major differences, measures are taken to improve the data.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

At present, there is no supply and use framework to investigate the statistical discrepancy between the production and the expenditure approach. This discrepancy is shown explicitly in the national accounts.

3.5 Revision studies

3.5.1 Studies and analyses of revisions are carried out routinely and used internally to inform statistical processes (see also 4.3.3)

For every revised version of the national accounts, revision studies are carried out to assess the sources of errors, omissions, and fluctuations in the data. The results of these studies are used to refine preliminary data and are generally referred to in the national accounts publications.

4. Serviceability

4.1 Periodicity and timeliness

4.1.1 Periodicity follows dissemination standards

Annual GDP estimates meet the General Data Dissemination Standards (GDDS) on periodicity.

4.1.2 Timeliness follows dissemination standards

Annual GDP estimates meet the GDDS requirement on timeliness.

4.2 Consistency

4.2.1 Statistics are consistent within the dataset

Although GDP estimates by activity and expenditure components are internally consistent from a conceptual point of view, there is a statistical discrepancy that is shown independently. This discrepancy is attributed to the expenditure approach. As a consequence of more comprehensive estimates of the unobserved economy, the statistical discrepancy has been diminishing during the past years (from 15 percent in 1998 to 3 percent in 2001). However, it is unclear to what extent the corrections for the errors in the estimates of the unobserved economy will influence the statistical discrepancy.

4.2.2 Statistics are consistent or reconcilable over a reasonable period of time

Annual national accounts data are available since 1991. In case of changes in source data, methodology, or statistical techniques, Goskomstat adjusts back series for at least 10 years. Unusual changes in economic trends are explained in detail in the publication of national accounts.

4.2.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

The national accounts data are largely consistent with government finance statistics. With BOP statistics, however, inconsistencies exist. The data that Goskomstat collects with regard to imports and exports include value information on bartered goods. These data are on a cash basis. Before using them in the compilation of BOP statistics, NBT adjusts these data to an accrual basis, while Goskomstat does not for the national accounts. Furthermore, NBT makes corrections for the expenses of residents abroad and of nonresidents in the national economy. These corrections are not incorporated in the national accounts. NBT also adjusts the valuation of imports from c.i.f. to f.o.b., while imports in national accounts are valued on a c.i.f. basis. It is recommended that Goskomstat takes these corrections into consideration when compiling the national accounts.

4.3 Revision policy and practice

4.3.1 Revisions follow a regular and transparent schedule

The revision cycle for annual national accounts is as follows:

  • preliminary estimates—after four months;

  • first revised estimates—after eleven months; and

  • final estimate—after fifteen months.

This revision policy is well established and generally stable from year to year. Users are informed of this revision policy through the annual SPS.

4.3.2 Preliminary and/or revised data are clearly identified

Preliminary and revised data are clearly identified in the national accounts publication by means of footnotes referring to the status of the published data.

4.3.3 Studies and analyses of revisions are made public (see also 3.5.1)

The reasons and the impact of revisions are generally analyzed. Upon request, users can obtain the results of these studies.

5. Accessibility

5.1 Data accessibility

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

The national accounts data are published clearly. The yearly national accounts publication contains charts and tables to facilitate analysis. Different levels of aggregation are presented, of which the most detailed contains 15 industries. For reasons of readability, only a limited numbers of years are shown. Upon request, however, longer time series can be obtained.

5.1.2 Dissemination media and format are adequate

National accounts statistics are disseminated in hard copy only. Several publications of the Goskomstat contain data on national accounts, although the annual national accounts publication is the most comprehensive.

Publications with national accounts statistics are readily available for a limited group of preferred users.

5.1.3 Statistics are released on a preannounced schedule

The annual catalog with the foreseen publications for the following year contains also information of their scheduled time of dissemination. Generally, this schedule is met.

5.1.4 Statistics are made available to all users at the same time

Users who apply for a copy of the publication in advance will receive the statistics simultaneously. The press is not briefed in advance.

5.1.5 Statistics not routinely disseminated are made available upon request

General statistics not disseminated, as well as customized tables, can be provided against payment. Goskomstat receives requests (and provides the data) for customized tables almost daily.

5.2 Metadata accessibility

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from internationally accepted standards, guidelines, or good practices are annotated

Only general notes on the compilation of national accounts are available and can be obtained upon request. Documentation on scope, classifications, and basis of recording are not available. Tajikistan has recently subscribed to the GDDS and is currently waiting for technical assistance to draft the necessary metadata.

5.2.2 Levels of detail are adapted to the needs of the intended audience

The annual catalog of publications of Goskomstat contains the details of prices and the venues of sales as well as an order form. Quarterly press announcements include name and contact details of persons in charge of different statistical areas at Goskomstat.

5.3 Assistance to users

5.3.1 Contact points for each subject field are publicized

All publications of Goskomstat hold contact details of persons in charge of the disseminated statistics. Quarterly press announcements include name and contact details of persons in charge of different statistical areas at Goskomstat. Inquiries can be made by mail or by phone.

5.3.2 Catalogs of publications, documents, and other services, including information on any changes, are widely available

Annual catalogs of publications are sent to a limited group of users. These catalogs include the price of all publications.

Table 1.

Republic of Tajikistan: Data Quality Assessment Framework (July 2003): Summary of Results for National Accounts Statistics

(Compiling Agency: Goskomstat)

article image
article image

II. Price Statistics (Consumer Price Index)

0. Prerequisites of quality

0.1 Legal and institutional environment

0.1.1 The responsibility for collecting, processing, and disseminating the statistics is clearly specified

The legal basis for the collection, compilation, and dissemination of the statistical information is the Law on State Statistics of the Republic of Tajikistan adopted in May 1997 and amended in April 2003 to reflect the new status of the Goskomstat as an independent state statistics committee. The law includes provisions for all statistical agencies of Tajikistan but assigns Goskomstat the leading role in production and dissemination of statistical information. The upgraded status of Goskomstat to a statistical committee, as well as its composition and responsibilities, were spelled out in the Government Decision of February 2001.

According to the law, an annual SPS is developed by the Goskomstat in collaboration with the other data-producing state agencies involved in the production of particular sets of statistics. This program includes, among other things, the schedule of surveys to be conducted by the Goskomstat or by other state agencies, the categories of administrative data to be provided to Goskomstat, as well as their schedules of reporting. The agreed draft program is submitted first to the Tajik government and subsequently to the advisory body—the Statistics Committee of the CIS for their revision and comments. The final version of the work program is then submitted to the president of Tajikistan for approval. Any additional requests received during the year are centralized and considered for the next year’s working plan.

0.1.2 Data-sharing and coordination among data-producing agencies are adequate

The SPS clearly defines the responsibilities of each state agency for the datasets to be provided to Goskomstat for statistical purposes at the appropriate frequency and timeliness. Data-sharing among statistics-producing agencies and the accuracy of data received is considered satisfactory by the Goskomstat’s managers and staff. The only evidence of duplication of efforts regards the price statistics, namely the production of a consumer price indicator4 by the NBT for internal purposes only.

0.1.3 Individual reporters’ data are to be kept confidential and used for statistical purposes only

The Law on State Statistics contains provisions guaranteeing the confidentiality of data provided to the Goskomstat and other authorized state agencies. Article 9 of this law states that “employees of state statistical bodies must protect the confidentiality of existing statistical information.” However, this provision is ambiguous and raises suspicions with regard to an even protection of confidentiality of both physical and legal persons giving leeway to interpretations. The law might benefit from a clear statement regarding the confidentiality of individual statistical information provided by legal persons.

The Goskomstat enforces confidentiality restrictions. All data relating to information classified as “confidential” (military industry, some of the mining operations, etc.) are transmitted by special courier to the special security section of Goskomstat. This section has the responsibility to (i) protect the confidentiality of certain categories of data and (ii) classify data as confidential. Any request for confidential information by department heads or from outside Goskomstat requires the authorization of the security section. For example, in the case of national accounts, special aggregations of data are made in order to protect the confidentiality of such data. Reportedly, there have not been any cases of violations of the confidentiality requirements. Confidentiality of data is appropriately guarded during storage, and the archives are under the responsibilities of a special commission. Staff are subject to disciplinary action for violation of confidentiality.

Respondents are informed in writing of their rights and obligations with regard to the provision of information. At the beginning (cover) of Goskomstat’s publications, a quotation from the law is included with reference to the use of information for statistical purposes solely.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

The Law on State Statistics provides for mandatory reporting and penalties for noncompliance. The enterprises that fail to report accurate and timely data are liable for fines. In practice, the sanctions are applied only in cases where the enterprises refuse to respond after receiving a warning. The regional offices have the right to impose fines in such cases, however small, but most often the enterprises choose to cooperate rather than to be liable to the board meetings deliberations of the responsible ministry.

0.2 Resources

0.2.1 Staff, facilities, computing resources, and financing are commensurate with statistical programs

Overall, the Goskomstat staff number is considered appropriate for the current duties; however, the number is considered insufficient in the regional offices.5 Currently, the total staff at Goskomstat headquarters consists of 265 employees, of which 166 are working in the computing center. The regional offices comprise 595 staff. Staff has been downsized6 over the past 10 years (about 70 percent), leading to the reorganization of the institution and restructuring of the work program. Currently, a certain degree of stability has been achieved.

Requests for staff increases have been addressed to the government at times without fruitful results. Owing to the low salary levels, which are not competitive with those of other agencies, it is difficult to retain younger qualified staff. However, there have been recorded cases of staff returning to Goskomstat in recognition of its good work environment.

For the price indices processes (CPI and PPI), there are five staff at headquarters with adequate qualification and experience in index tabulation (with a university degree) and household survey, five staff in the computing center with responsibilities for data collection and processing (two of which have a university degree), and about ten price collectors spread among the three big regions. Currently, at Goskomstat, there are three vacancies in the CPI unit and another three in the price computing section. This situation may be explained by a certain inconsistency between the level of qualification required for these positions and the salaries offered at Goskomstat. Goskomstat’s price collector number is far from sufficient for the workload (three to four additional staff would be needed). A regular internal training for staff is organized, and two to three persons have participated in international or regional courses. The head of the Price Division, who has years of experience in the field, attended the IMF seminars in Vienna (2002) and Kiev (1998), and found them extremely valuable. He also attends, when possible, regional meetings on price comparison and is very interested in implementing new developments in price statistics.

Computing resources are below needs, particularly in regional centers. The tabulation of the CPI and PPI is done on two computers (plus one in the head’s office, mainly used for analysis) of good capacity (Pentium IV). The upload of average price data is done by the price computing center, equipped with two computers only. Software for the CPI and PPI processing is in need of upgrade. There is scope for improvement in the process of data transfer from regions. No region provides data in electronic format.

Budget allocation for Goskomstat provided from the government budget is rather tight for the work program. This budget is not sufficient to support development work or computing enhancements. Local offices are supported by the local budget. Financial assistance is sought in order to permit statistical development.

0.2.2 Measures to ensure efficient use of resources are implemented

Analyzing the work program each year, management reviews priorities in developing the program. Efficiencies are sought throughout the work process at Goskomstat. Effective collaboration is sought with the other state agencies for the development of the SPS that defines, among other things, the flow of information at the national level to avoid duplication. Reviews of staff performance are conducted with variable frequency depending upon the position (once in two to three years).

Efforts are made to provide staff in regional offices, through periodic training, with clear directions regarding methods and procedures applicable for specific work. Inspections are made regularly (yearly) to check the accuracy of data collection and compilation processes. New possibilities could be explored, nevertheless, for more efficient data transmission mechanisms, use of consistent concepts, and classification across datasets (e.g., use the same classification for households final consumption in national accounts and HES, etc.)

In view of the limited resources, tasks are reassigned as needed to even the workload. When one person is absent for various reasons, his or her work is completed by the rest of the team. During peak periods, staff work overtime to ensure the quality of data and to meet the scheduled deadlines.

Expert assistance from international organizations to evaluate statistical methodologies and compilation methods is considered as very beneficial.

0.3 Relevance

0.3.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored

Users are kept informed on the current work program, including through the annual catalog of publications, and their feedback is sought. However, the concept of users is limited to a few categories (so-called “principal users”), including government agencies, universities, embassies, and most important business companies. When the principal users receive the catalog of publications, they are also invited to submit requests for publications and types of information of interest and to provide comments as needed in writing. Users’ comments are monitored and responses are provided. Although there is no established schedule for meetings with users, the Goskomstat participates in meetings organized by the government or various ministries, where media and various users are present. Goskomstat is also invited to similar regional meetings organized by the local government units.

Goskomstat participates in statistical meetings and seminars organized by international or regional organizations such as the CIS, IMF, ILO, and UN.

0.4 Other quality management

0.4.1 Processes are in place to focus on quality

Within the Goskomstat, there is recognition that quality is a pillar of sound statistical work, and management is sensitive to all dimensions of quality. The annual SPS is monitored on an ongoing basis and improvements are sought in all areas. A methodological commission composed mostly of heads of departments (16–23 persons) meets regularly to discuss all proposals for methodological improvements and new developments.

Recent initiatives on quality improvement are the data module ROSC and the country’s decision to participate in the IMF’s GDDS.

0.4.2 Processes are in place to monitor the quality of the statistical program

A new section dealing with international methods and standards has been set up recently at Goskomstat in view of the country’s participation in GDDS. The quality control of each employee’s work is tightly monitored within each department. Staff are made aware of the importance of quality through internal training programs.

0.4.3 Processes are in place to deal with quality considerations in planning the statistical program

The Goskomstat monitors the quality of its work programs with regard to trade-offs between improvements and available resources. It periodically reviews survey results to identify shortcomings. It considers feedback from users on newly emerging data requirements to the extent possible in planning the statistical program. However, in the case of national accounts and prices, most methodological improvements are the result of the technical assistance provided by international organizations. More internal initiatives for improvement would be very beneficial.

1. Assurances of integrity

1.1 Professionalism

1.1.1 Statistics are produced on an impartial basis

Article 4 of the Law on State Statistics prescribes that the Goskomstat and the territorial statistical services under its control are the official bodies of state statistics and that they gather, monitor, process, analyze, aggregate, disseminate, and store statistical information in accordance with this law.

Staff is hired on the basis of competence. Vacancies are published in the press and a commission of senior staff of the Goskomstat investigates all the application forms in order to identify the best qualified candidate, after which this commission proposes to the chairman to hire that candidate. Promotions of staff are based on personal performance and proficiency. Moreover, the work environment at Goskomstat supports a culture of competence.

The chairman of the Goskomstat is appointed or dismissed by the president of the republic according to Article 69 of the Constitution, which regulates appointment and dismissal of heads of government bodies.

Staff can participate, to a certain extent, in seminars or courses organized by international organizations.

All publications are reviewed by a commission of senior staff in order to guarantee consistency between statistics.

1.1.2 Choices of sources and statistical techniques, as well as decisions about dissemination, are informed solely by statistical considerations

The Goskomstat is independent in its choice of appropriate source data and statistical techniques. In its publications, the Goskomstat attempts to present the statistics in a way that users understand and appreciate. Statistical data that are considered to be secret are disseminated at an aggregated level with other data categories (e.g., statistics on defense or certain mining operations), although they are not confidential from a statistical point of view. Even though this practice is used to protect confidentiality of individual data, in the above case, it is not recommended.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The Goskomstat is entitled to comment and respond on any misused or misinterpreted statistics. When new statistics are published, the Goskomstat explains to the government and other users that the statistics are the result of a process based on international guidelines with regard to the collection and compilation rules used.

1.2 Transparency

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The publications and surveys of the Goskomstat contain references and relevant articles of the Law on State Statistics under which the statistics are collected, processed, and disseminated. Users need previous consent from the Goskomstat to redistribute the statistics contained in its publications. This is mentioned in all publications.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

The government has access to the statistics prior to their release. However, this is not publicly notified since it is “presumed” to be known by the public.

1.2.3 Products of statistical agencies/units are clearly identified as such

All the publications of the Goskomstat are clearly identified by name and logo. When there are joint publications, the price data are clearly identified as products of the Goskomstat.

1.2.4 Advanced notice is given of major changes in methodology, source data, and statistical techniques

Changes in methodology, source data, or statistical techniques are announced in the price statistics publications at the time they are introduced. Any methodological changes have to be included in the SPS, and the public is notified in the publications.

1.3 Ethical standards

1.3.1 Guidelines for staff behavior are in place and are well known to the staff

The behavior of staff is governed by the Civil Code. Although there is no formal code of conduct, staff is, upon engagement, informed by management of the rules and guidelines, which they have to adopt. Reportedly, top management periodically reminds staff of these rules and guidelines during meetings.

2. Methodological soundness

2.1 Concepts and definitions

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

The concepts and definition for household consumption expenditures and determination of the index weights follow broadly the ILO guidelines and the CIS regional Statistics Committee recommendations. The IMF technical assistance mission of 1995 assisted with the development of the CPI. However, in keeping with the report requirements of the CIS committee, the adopted classification serves well for regional comparison purposes but to a lesser extent the internationally comparability. Moreover, the selection of expenditure items to be included in the CPI basket follows the 0.1 percent threshold. This affects the imputed rentals of owner-occupied dwellings, among others, which are excluded for this reason. An exercise needs to be undertaken to recalculate an imputed value.

The index reflects the price change of goods and services consumed by the households. Expenditure weights are mainly derived from the annual HES. Prices collected by the monthly survey relate, as far as possible, to goods and services well specified in terms of product and transaction characteristics. The CPI basket, dominated by food products (over 70 percent), reflects a typical transition economy pattern.

Expenditure estimates for both commodities and services are produced in sufficient detail to permit analysis at various levels of detail7 of the country classification system.

2.2 Scope

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices

The weights used for the CPI market basket are derived from the expenditure data from the annual HES and adjusted by the sales data structure from the annual retail trade survey (RTS). The HES is representative for all types of resident households and income levels for both urban and rural8 areas. Households whose major activities are small businesses or farming are also included. However, the representativeness of the high-income population is weaker than for the other categories.

Transactions considered for the CPI correspond generally to purchases of goods and services for consumption. Expenditures on fixed capital formation, such as dwellings or cattle for agriculture, are excluded. However, payments in kind by employers, although considered insignificant, are included within the CPI scope, while the general practice does not consider them.

The item coverage in the CPI expenditure aggregate includes all market goods and services purchased by the population for consumption purposes and own-account production of goods for own final consumption (mainly agricultural products9). Imputed rentals of owner occupied dwellings are excluded because they did not exceed the threshold used according to the CIS methodology. The CPI Manual recommends that the imputed rentals of owner-occupied dwellings be included in the index compilation irrespective of their low share. The estimate produced by the national accounts could be used for this purpose. The resulting weight will be added to the rental weight and moved forward by the change in rental values.10

Luxury products and some durables (e.g., cars) are also missing from the CPI scope for various reasons. Excluded as recommended by standards are expenditures on illegal market goods and services, gifts, and donations (e.g., wedding gifts), and direct taxes.

2.3 Classification/sectorization

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices

Institutional units and transactions follow broadly the 1993 SNA standards. Consumption items and the index aggregation are based on a regional classification developed for the CIS countries for purposes of regional comparison of commodities recorded by retail trade. The aggregation categories, particularly with regard to services, are different from the COICOP, which is the recommended international standard. A national classification following COICOP is under development. CPC is not used as to date.

2.4 Basis for recording

2.4.1 Market prices are used to value flows and stocks

Both basket weights and monthly prices used in the compilation of the CPI are based on market prices (purchasers’ prices), including taxes and trade and transportation margins. Purchasers’ prices are also imputed for the own-account production of goods for own final consumption based on observed market prices. For certain products, mainly those sold in the informal markets, the price collectors obtain the actual market transaction prices by bargaining with vendors. Detailed characteristic information, including terms of transactions, is included as part of the product specifications for monthly price collection, following the previous IMF mission recommendations.

2.4.2 Recording is done on an accrual basis

Both the expenditure data from the HES and the price data for the monthly price survey are recorded on an accrual basis. Prices are collected around the period 20–25th of the reference month for foods stuffs and during the 15–25th for the other categories. Tariff prices are collected monthly by phone on a routine checking basis. When tariff changes occur, a notification in writing is sent to the Goskomstat.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices

Secondhand goods or insurance services are not included in the current CPI, implying that there are no transactions that require calculation of net weights. The HES does not collect information regarding the acquisitions and sales of used cars. Most of the secondhand cars are imported from Russia and might have an important share in the personal transportation vehicle purchase. It would be useful to include a question in the HES questionnaire to allow for used cars expenditures. If this were found to be significant, the product should be included in the market basket. The estimate of the weights for automobile purchases will be compiled as acquisitions less sales.

3. Accuracy and reliability

3.1 Source data

3.1.1 Source data are obtained from comprehensive data collection programs that take into account country-specific conditions

The basic structure of the weights for the CPI market basket is derived from the annual expenditure data from the HES. These weights are adjusted by the sales data from the six-month RTS. The sample frame for the HES is the population census (the last one was organized in 2000).

The HES is conducted on a permanent basis using a monthly sample of 1,000 households rotated annually. The survey is designed to ensure coverage across the country. However, the sample seems rather small.11 A multistage stratified sampling is used, structured by rural and urban area, by oblast and by types. A random selection of households within the strata is applied to ensure representativeness of all categories of households by size and income. The survey is revised periodically (every three to five years) to take account of changed circumstances. The regional offices ensure good cooperation of households, and the response rate is practically 100 percent. In cases where families selected in the sample have moved, a replacement in the same category is sought, to maintain the sample composition. Overall, the replacement rate is no more than 10 percent annually.

The HES expenditure data are supplemented by data from the RTS. These data are used to provide additional item detail and reliable information to cross-check expenditure data collected by the HES in order to make adjustments to the CPI weights. The RTS is based on a purposive selection of outlets (more than 8,000 vendors, most of which are nonstate entities) and products (about 92). The response rate is practically 100 percent, and the results are extrapolated based on population.

A regular monthly price survey is conducted by personal visit to the selected outlets that include supermarkets, department stores, open markets, retail shops, street vendors, etc., spread across the country (mainly in cities or small towns serving also the rural areas). A stratified selection of outlets by type, location, and size is applied based on information from the RTS. Overall, the number of outlets in the three oblasts12 is over 600. This sample is updated annually to include new emerging outlets or to replace those that disappear.

In principle, all representative products for the household consumption are covered, including those sold in parallel markets. The CPI basket includes 274 representative items for which prices are collected monthly. For each item, five to six prices are collected in each oblast.

For the most part, the survey questionnaires for the HES and price collection are adequately designed to ensure an efficient collection of data needed for CPI compilation. However, further improvements are needed to ensure a good representativeness of all items. For example, certain items13 for which expenditure data are collected in the HES are not represented in the CPI, while others (alcoholic beverages) are underrepresented. The rental services are below 2 percent, and alcohol and tobacco categories together represent less than 1 percent. Notwithstanding the country’s specific circumstances and the recent period of instability, these figures appear to be significantly underestimated.

While the weights have been revised annually in the last few years (from 2,000 onwards), the product basket composition and number have recorded little change in the past 10 years. Progress has been registered insofar as an important additional number of products are priced to serve as a basis for analysis for product replacement or eventual introduction of new products.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

In general, source data are broadly consistent with the needs of the CPI program. The HES and CPI are closely coordinated, and both sets of statistics are integrated in the same unit. However, the HES sampling could be further improved to better serve the CPI needs. For example, a question on the estimation of imputed rent for owner occupants could be included in the form, and the results adjusted on the basis of complementary information from other surveys and censuses. Similar adjustments should be made to those categories of expenditures considered insignificant, if the results are judged unrealistic.

3.1.3 Source data are timely

The frequency of data collection supports the publication of the CPI no later than the 12th of the following month. Both monthly prices and expenditure data needed for the estimation of CPI are collected and processed on a timely basis following a fixed schedule.

3.2 Assessment of source data

3.2.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide statistical processes

The results of both HES and price survey are checked for consistency and extreme values. Statistical analysis is carried out to determine the sample variation for all tabulated values, but these results were not made available for the mission. CPI monthly price data are usually monitored closely for extreme values or unusually large movements in item indices. These are traced back to the reporting retail outlet if necessary and confirmed. Verified prices and elementary indices are included in the index or corrected if found to be in error. There are no automatic procedures for detecting unusually large movements in prices from the index.

Inspection visits in the regions are also organized periodically (once a year) by the Goskomstat, and the price collectors are dismissed if an inaccuracy or same error occurs the second time. However, the limited number of price collectors (10 for the whole country), combined with insufficient computing resources in the regions, raises doubts about the accuracy of data collection.

Nonresponse is not an issue for any of the surveys. Individual price movements are compared with external data, such as the weekly collection of 25 staple products (mainly volatile food products), retail trade data, etc.

3.3 Statistical techniques

3.3.1 Data compilation employs sound statistical techniques to deal with data sources

The staff makes efforts to ensure a high quality of collected and processed data to the extent possible. However, the accuracy check is mostly done manually before and after the upload of individual data, owing to the unsophisticated software available, and it depends greatly on the experience and judgment of staff.

Temporarily missing prices are imputed using appropriate techniques. If a product is missing, the price for the same product from the other outlets is imputed. Large discrepancies in price change across the country, however, have not been noticed (at most 10–20 percent). If the product is missing from all outlets, the price of another product with similar characteristics is imputed, or the price based on the average change of a group of similar products. Alternatively, the last reported price is carried forward if the group evolution is small.

For missing seasonal items (fruits and vegetables), prices are only collected during the season. During off season, either the products are dropped or the last reported price is carried forward until the product reappears in the market and collection resumes. Because of the high share of food stuffs in the CPI (over 70 percent), the index fluctuations that are due to seasonal products (around 2–5 percent) and price variation by season of other foodstuffs can be noticeable in the CPI.

Prices for products that become permanently unavailable during the year are held constant until the beginning of the next year when the sample of outlets and specifications is updated. Prices for new types of product specifications that are expected to be used in the index for the following year are collected but are not used in the index. These prices are used for estimating the base-year price for the following year. New products are introduced into the index only during the weight revisions. Quality adjustments are rarely made and then by using simple techniques (usually by using overlap prices method). It should be noted that prices that are held constant for a long period during the year have the potential of introducing bias into the index.

3.3.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

The CPI weights are based on household expenditure, largely following the 1993 SNA with the main exception of the classification used. The expenditure data in HES cover all types of expenditures by the surveyed households including those in parallel markets. The weights reflect the economic cost to consumers of goods and services in the period in which they were purchased, regardless of how the expenditures were financed, whether by cash purchase or by credit.

Elementary price aggregates are derived using the arithmetic mean of short-term price relative formula. Beginning with 1994, as recommended by the IMF mission, the modified Laspeyres formula is used to aggregate higher-level indices. In the following years, the weights of 1993 had been held constant. After 2000, the weighting base period has changed annually, and the household expenditures of the first nine months of the previous year have been used to derive the index’s weights. These weights are corrected as explained previously with the retail trade data. Annual results of the HES usually become available at the end of the first quarter of the following year. The weights derived from the annual results are compared with those of the nine months, and they could be used to replace the latter if the differences are significant.14 So far, this exercise has not been found necessary. Plans are to use both weights (nine months and one year) for the 2005 CPI as an exercise to make a better comparative analysis at the detailed level. On the basis of results and taking into account the inflation evolution, a decision will be taken whether it is reliable or not to use 2003 as a fixed-base year for a five-year period.

Tajikistan does not produce regional CPIs, except implicitly to compile the national CPI. The index is compiled directly on the basis of elementary index data transmitted by the three regions (oblasts) and using their population as weights. Elementary data are transmitted on hard copy and processed in the computing center at the Goskomstat. Conceptually, the expenditure shares of the oblasts in total expenditures derived from the HES should be used as weights to better account for expenditure cost movement. This is obvious in this case where Dushanbe, which has the highest income population among the three oblasts, scores the lowest because of the smallest population share (27 percent).

3.4 Assessment and validation of intermediate data and statistical outputs

3.4.1 Intermediate results are validated against other information where applicable

Monthly CPI series are compared for consistency with commensurate data from the PPI, import/export price indices, and other relevant price data (e.g., retail price data).

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated

Unusual index movements are investigated to determine their causes. In one case it was determined that the prices for bread were unusually higher as compared to those of the PPI, which were explained by the increase in the transportation cost occurred during the month. In other cases, differences in prices were the result of the exchange rate movement, etc.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

Since only national-level price indices are compiled, there is no discrepancy arising from imputation inconsistencies between aggregations by product or by geographical area.

3.5 Revision studies

3.5.1 Studies and analyses of revisions are carried out routinely and used internally to inform statistical processes (see also 4.3.3)

The CPI is final when first released, and the index is not subsequently revised. When methodology is revised, the index series are not revised backwards.

A regular analysis of the nine-month and one-year weights is done annually to assess eventual significant changes in the consumption pattern that may entail the use of the annual weights. The results of HES are also compared with corresponding data from other sources.

4. Serviceability

4.1 Periodicity and timeliness

4.1.1 Periodicity follows dissemination standards

The CPI is compiled and published monthly, thus meeting GDDS standards.

4.1.2 Timeliness follows dissemination standards

The CPI is published no later than the 12th of the following month, thus exceeding the GDDS standards.

4.2 Consistency

4.2.1 Statistics are consistent within the dataset

The index is internally consistent since there is a single aggregation at the national level.

4.2.2 Statistics are consistent or reconcilable over a reasonable period of time

Consistent CPI series are available monthly back to 1994. Unusual changes in economic trends are explained in the analytical text included in publications.

4.2.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

The CPI is consistent with the other price indices, such as the PPI, agricultural price index, import/export price index, etc. However, the consistency with the national accounts is not as straightforward owing to the different classifications employed. While national accounts produce the household final consumption according to the COICOP, the HES, and CPI are still based on the regional classification developed by the CIS.

4.3 Revision policy and practice

4.3.1 Revisions follow a regular and transparent schedule

The CPI weights are revised regularly (annually for the last few years). This revision schedule is generally known by the main users.

4.3.2 Preliminary and/or revised data are clearly identified

The index number is considered final when first published. The publication does not mention the weights used for the CPI or their frequency of revision.

4.3.3 Studies and analyses of revisions are made public (see also 3.5.1)

The results of the revision studies, such as weights change, for the CPI are not published. Moreover, published documentation for the CPI does not include any direct reference to the HES, as the weighting base, or to the impact of revisions.

5. Accessibility

5.1 Data accessibility

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

The CPI publication contains clear tables and charts, as well as detailed analyses of data. A reasonable amount of detail is published. The index is presented as the change from the last month and from the same month of the previous year. It also appears as a quarterly or yearly cumulated change.

5.1.2 Dissemination media and format are adequate

In addition to the monthly publication in the bulletin of social indicators, CPI series are published in a three-year publication on prices, as well as in the yearly book. It should be noted that the cost of publications is discouraging for the public at large. Moreover, requests from individuals are monitored, along with the subscribers, who are questioned about their reasons to buy these publications. No Internet dissemination is available yet.

5.1.3 Statistics are released on a preannounced schedule

The CPI is disseminated according to a preannounced monthly calendar that can be found in the special catalog of information. This catalog presents the periodicity and the scheduled timeliness for Goskomstat’s publications.

5.1.4 Statistics are made available to all users at the same time

In principle, the CPI is released to all users simultaneously.

5.1.5 Statistics not routinely disseminated are made available upon request

Nonconfidential data at a higher level of detail are available upon request and for a fee. However, this fact is not widely known, and a preferential treatment seems to be applied for main users.

5.2 Metadata accessibility

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from internationally accepted standards, guidelines, or good practices are annotated

The CPI metadata available in the monthly publication are very brief. More details are provided in the three-year publication on price statistics but without any information about shortcomings in the data.

5.2.2 Levels of detail are adapted to the needs of the intended audience

Although a short documentation is available in various publications, a comprehensive document describing the concepts, sources, and methods used for the collection and compilation of the CPI is not available. In order to better document details for internal and external users, it is recommended to produce and publish such a methodology. This methodology should be updated when methodological changes occur.

5.3 Assistance to users

5.3.1 Contact points for each subject field are publicized

All statistical products of the Goskomstat include a contact person, telephone number, mailing address, and e-mail address of the price unit. A prompt service is provided to inquirers.

5.3.2 Catalogs of publications, documents, and other services, including information on any changes, are widely available

A catalog of publications updated annually is available for consultation but not widely distributed. The main users, who are also subscribers, receive a copy of this catalog annually.

Table 2.

Republic of Tajikistan: Data Quality Assessment Framework (July 2003): Summary of Results for Consumer Price Index

(Compiling Agency: State Committee on Statistics)

article image
article image

III. Price Statistics (Producer Price Index)

0. Prerequisites of quality

0.1 Legal and institutional environment

0.1.1 The responsibility for collecting, processing, and disseminating the statistics is clearly specified

The legal basis for the collection, compilation, and dissemination of the statistical information is the Law on State Statistics of the Republic of Tajikistan. It was adopted in May 1997 and amended in April 2003 to reflect the new status of the Goskomstat as an independent state statistics committee. The law includes provisions for all statistical agencies of Tajikistan but assigns Goskomstat the leading role in production and dissemination of statistical information. The upgraded status of Goskomstat to a statistical committee as well as its composition and responsibilities were spelled out in the Government Decision of February 2001.

According to the law, an annual SPS is developed by the Goskomstat in collaboration with the other state agencies involved in the production of particular sets of statistics. This program includes, among other things, the schedule of surveys to be conducted by the Goskomstat or by other state agencies, the categories of administrative data to be provided to Goskomstat, as well as their schedules of reporting. The agreed draft program is submitted first to the Tajik government and subsequently to the CIS for their revision and comments. The final version of the work program is then submitted to the president of Tajikistan for approval. Any additional requests received during the year are centralized and considered for the next year’s working plan.

0.1.2 Data-sharing and coordination among data-producing agencies are adequate

The SPS clearly defines the responsibilities of each state agency for the datasets to be provided to Goskomstat for statistical purposes at the appropriate frequency and timeliness. Data-sharing among statistics-producing agencies and the accuracy of data received is considered satisfactory by the Goskomstat’s managers and staff. The only evidence of duplication of efforts regards the price statistics, namely the production of a consumer price indicator15 by the National Bank of Tajikistan (NBT) for internal purposes only.

0.1.3 Individual reporters’ data are to be kept confidential and used for statistical purposes only

The Law on State Statistics contains provisions guaranteeing the confidentiality of data provided to the Goskomstat and other authorized state agencies. Article 9 of this law states that “employees of state statistical bodies must protect the confidentiality of existing statistical information.” However, this provision is ambiguous and raises suspicions with regard to an even protection of confidentiality of both physical and legal persons. The law might benefit from a clear statement regarding the confidentiality of individual statistical information provided by legal persons.

The Goskomstat enforces confidentiality restrictions. All data relating to information classified as “confidential” (military industry, some of the mining operations, etc.) are transmitted by special courier to the special security section of Goskomstat. This section has the responsibility to (1) protect the confidentiality of certain categories of data and (2) classify data as confidential. Any request for confidential information by department heads or from outside Goskomstat requires the authorization of the security section. For example, in the case of national accounts, special aggregations of data are made in order to protect the confidentiality of such data. Reportedly, there have not been any cases of violations of the confidentiality requirements. Confidentiality of data is appropriately guarded during storage, and the archives are under the responsibilities of a special commission. Staff are subject to disciplinary action for violation of confidentiality.

Respondents are informed in writing of their rights and obligations with regard to the provision of information. At the beginning (cover) of Goskomstat’s publications a quotation from the law is included with reference to the use of information for statistical purposes solely.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

The Law on State Statistics provides for mandatory reporting and penalties for noncompliance. The enterprises that fail to report accurate and timely data are liable for fines. In practice, the sanctions are applied only in cases where the enterprises refuse to respond after receiving a warning. The regional offices have the right to impose fines in such cases, however small, but most often the enterprises choose to cooperate rather than to be liable to the board-meeting deliberations of the responsible ministry.

0.2 Resources

0.2.1 Staff, facilities, computing resources, and financing are commensurate with statistical programs

Overall the Goskomstat staff number is considered appropriate for the current duties; however, the number is considered insufficient in the regional offices.16 Currently, the total staff at Goskomstat headquarters consists of 265 employees, of which 166 are working in the computing center. The regional offices comprise 595 staff. Staff has been downsized17 over the past 10 years (about 70 percent) leading to the reorganization of the institution and restructuring of the work program. Currently, a certain degree of stability has been achieved.

Requests for staff increases have been addressed to government at times without fruitful results. Owing to the low salary levels, which are not competitive with those of other agencies, it is difficult to retain younger qualified staff. However, there have been recorded cases of staff returning to Goskomstat in recognition of its good work environment.

For the price indices processes (CPI and PPI), there are five staff with adequate qualification and experience in index tabulation (with a university degree) and household survey, five staff in the computing center with responsibilities for data collection and processing (two of which have a university degree), and about 10 price collectors spread among the three big regions. Currently, there are three vacancies in the CPI unit and another three in the price computing section at Goskomstat. This situation may be explained by a certain inconsistency between the level of qualification required for these positions and the salaries offered at Goskomstat. The price collector number is far from sufficient for the workload (three to four additional staff would be needed). A regular internal training for staff is organized, and two to three persons have participated in international or regional courses. The head of the Price Division, who has years of experience in the field, attended the IMF seminars in Vienna (2002) and Kiev (1998) and found them extremely valuable. He is also attending, when possible, regional meetings on price comparison and is very interested in implementing new developments in price statistics.

Computing resources are below needs, particularly in regional centers. The tabulation of the CPI and PPI is done on two computers (plus one in the head’s office, mainly used for analysis) of good capacity (Pentium IV). The upload of average price data is done by the price computing center, equipped with two computers only. Software for the CPI and PPI processing is in need of upgrade. There is scope for improvement in the process of data transfer from regions. Only one region provides data in electronic format.

Budget allocation for Goskomstat provided from the government budget is rather tight for the work program. This budget is not sufficient to support development work or computing enhancements. Local offices are supported by the local budget. Financial assistance is sought in order to permit statistical development.

0.2.2 Measures to ensure efficient use of resources are implemented

Each year, management analyzes and reviews priorities in its work program. Efficiencies are sought throughout the work process at Goskomstat. Effective collaboration is sought with the other state agencies for the development of the SPS that defines, among other things, the flow of information at the national level to avoid duplication. Reviews of staff performance are conducted with variable frequency depending upon the position (once every 2 to 3 years).

Efforts are made to provide staff in regional offices, through periodic training, with clear directions regarding methods and procedures applicable for specific work. Inspections are made regularly (yearly) to check the accuracy of data collection and compilation processes. New possibilities could be explored, nevertheless, for more efficient data transmission mechanisms, use of consistent concepts, and classification across datasets (e.g., use the same classification for household final consumption in national accounts and HES, etc.)

In view of the limited resources, tasks are reassigned as needed to even the workload. When one person is absent for various reasons, his or her work is completed by the rest of the team. During peak periods, staff work overtime to ensure the quality of data and to meet the scheduled deadlines.

Expert assistance from international organizations to evaluate statistical methodologies and compilation methods is considered as very beneficial.

0.3 Relevance

0.3.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored

Users are kept informed on the current work program, including through the annual catalog of publications, and their feedback is sought. However, the concept of users is limited to a few categories (so-called “principal users”), including government agencies, universities, embassies, and most important business companies. When the principal users receive the catalog of publications, they are also invited to submit requests for publications and types of information of interest and to provide comments as needed in writing. Users’ comments are monitored and responses are provided. Although there is not an established schedule for meetings with users, the Goskomstat participates in meetings organized by government or various ministries, where media and various users are present. Goskomstat is also invited to similar regional meetings organized by the local government units.

Goskomstat participates in statistical meetings and seminars organized by international or regional organizations such as the CIS, IMF, ILO, and UN.

0.4 Other quality management

0.4.1 Processes are in place to focus on quality

Within the Goskomstat, there is recognition that quality is a pillar of sound statistical work, and management is sensitive to all dimensions of quality. The annual SPS is monitored on an ongoing basis and improvements are sought in all areas. A methodological commission composed mostly of heads of departments (16–23 persons) meets regularly to discuss all proposals for methodological improvements and new developments.

Recent initiatives on quality improvement are the data module ROSC and the country’s decision to participate in the IMF’s GDDS.

0.4.2 Processes are in place to monitor the quality of the statistical program

A new section dealing with international methods and standards has been set up recently at Goskomstat in view of the country’s participation in GDDS. The quality control of each employee’s work is tightly monitored within each department. Staff are made aware of the importance of quality through internal training programs.

0.4.3 Processes are in place to deal with quality considerations in planning the statistical program

The Goskomstat monitors the quality of its work programs with regard to trade-offs between improvements and available resources. It periodically reviews survey results to identify shortcomings. It considers feedback from users on newly emerging data requirements to the extent possible in planning the statistical program. However, in the case of national accounts and prices, most methodological improvements are the result of the technical assistance provided by international organizations. More internal initiatives for improvement would be very beneficial.

1. Assurances of integrity

1.1 Professionalism

1.1.1 Statistics are produced on an impartial basis

Article 4 of the Law on State Statistics prescribes that the Goskomstat and the territorial statistical services under its control are the official bodies of state statistics and that they gather, monitor, process, analyze, aggregate, disseminate and store statistical information in accordance with this Law.

Staff is hired on the basis of competence. Vacancies are published in the press, and a commission of senior staff of the Goskomstat investigates all the application forms in order to identify the best qualified candidate after which this commission proposes to the chairman to hire that candidate. Promotions of staff are based on personal performance and proficiency. Moreover, the work environment at Goskomstat supports a culture of competence.

The chairman of the Goskomstat is appointed or dismissed by the president of the republic according to Article 69 of the Constitution, which regulates appointment and dismissal of heads of government bodies.

Staff can participate, to a certain extent, in seminars or courses organized by international organizations.

All publications are reviewed by a commission of senior staff in order to guarantee consistency between statistics.

1.1.2 Choices of sources and statistical techniques as well as decisions about dissemination are informed solely by statistical considerations

The Goskomstat is independent in its choice of appropriate source data and statistical techniques. In its publications, the Goskomstat attempts to present the statistics in a way that users understand and appreciate.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

The Goskomstat is entitled to comment and respond on any misused or misinterpreted statistics. When new statistics are published, the Goskomstat explains to the government and other users that the statistics are the result of a process based on international guidelines with regard to the collection and compilation rules used.

1.2 Transparency

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The publications and surveys of the Goskomstat contain references and relevant articles of the Law on State Statistics under which the statistics are collected, processed and disseminated. Users need previous consent from the Goskomstat to redistribute the statistics contained in its publications. This is mentioned in all publications.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

The government has access to the statistics prior to their release. However, this is not publicly notified since it is “presumed” to be known by the public.

1.2.3 Products of statistical agencies/units are clearly identified as such

All the publications of the Goskomstat are clearly identified by name and logo. When there are joint publications, the price data are clearly identified as products of the Goskomstat.

1.2.4 Advanced notice is given of major changes in methodology, source data, and statistical techniques

Changes in methodology, source data or statistical techniques are announced in the price statistics publications at the time they are introduced. Any methodological changes have to be included in the SPS, and the public is notified in the publications.

1.3 Ethical standards

1.3.1 Guidelines for staff behavior are in place and are well known to the staff

The behavior of staff is governed by the Civil Code. Although there is no formal code of conduct, staff is, upon engagement, informed by the management of the rules and guidelines which they have to adopt. Reportedly, top management periodically reminds staff of these rules and guidelines during meetings.

2. Methodological soundness

2.1 Concepts and definitions

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

The concepts and definition of the PPI present certain deviations from the 1993 SNA and the PPI Manual in respect to weights determination, classification, and time of recording. One of the main departures from the 1993 SNA definition of market output concerns the use of total production realized, irrespective of destination.18 It is suspected that the valuation of the annual production used to derive the weight is not based on a unit value price that takes into account both export and domestic prices. A revision of the reporting form to include the value of domestic sales, exports, change in output inventories, and price units used might be useful.

Furthermore, the use of a domestic price for both the products for export and those for the domestic market has the potential of introducing serious bias in the index if the monthly evolution and trend of the two prices are different (see 3.3.2).

The index is intended to measure price changes in products produced by industrial activities. Weights are annually updated and are derived from the annual production values of the industrial activities. PPI estimates are compiled by industry and by commodity.

Prices recorded by the monthly survey are considered ex-factory gate prices, net of VAT or excise taxes, and trade and transportation margins for the products sold on the domestic market19 only. As follow-up to the IMF mission of 1994, product specifications for pricing have been improved lately and cover both product characteristics and transaction characteristics.

2.2 Scope

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices

The PPI covers manufacturing, electricity, and gas activities, which account for about 40 percent of the total output of the economy. All resident enterprises for the defined sectors are in scope. For the sectors covered, changes in inventories, work-in-progress, and illegal market goods are not accounted explicitly, since the output valuation adopted by Tajikistan is total production.

2.3 Classification/sectorization

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices

The industrial activities are classified at four-to-five digit level of the old Soviet Union classification—OKONKH (All Union Classification of Industrial Activities), which differs from the internationally used classifications with respect to division grouping. Plans are to implement ISIC starting with next year. Products are classified of four–five digit level of the CPC.

2.4 Basis for recording

2.4.1 Market prices are used to value flows and stocks

Output used to derive the weights refers to the total production of the activities in the index scope, with departures from the market output valuation as noted in 2.1.1. Prices recorded by the monthly survey approximate ex-factory gate prices, net of VAT or excise taxes, and trade and transportation margins for the products sold on the domestic market only. The same prices are used for exported goods as well, where applicable, which is not a recommended practice.

2.4.2 Recording is done on an accrual basis

Prices reported monthly are average prices recorded during the period 1–25th of the current month, as noted in the product specifications. However, the principle of recording output on an accrual basis is not respected, since production valuation is not compiled on the basis of sales. All production is considered output even when shipments do not occur. Moreover, adjustments to the recording time are mandatory because of the existence of barter transactions.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices

As the PPI is a gross sector index, and no stage of production index is compiled, transactions between enterprises are recorded on a gross basis, and no netting procedures are needed.

3. Accuracy and reliability

3.1 Source data

3.1.1 Source data are obtained from comprehensive data collection programs that take into account country-specific conditions

The weights used for the PPI compilation are derived from the annual production values of the industrial activities of the last (but one) year. The enterprises submit to Goskomstat monthly, quarterly and annual reports of various indicator coverage and level of detail. The information about the enterprises is maintained in a state register, which is periodically updated (usually annually) based on the regular information received from different sources. Information about legal persons (unincorporated enterprises mostly belonging to households) is received from the MOF.

The current PPI uses the weights of 2002. In recent years, the weights have been updated annually to account for changes in product and activity evolution. However, there have not been recorded significant changes from one year to another.

A purposive selection of enterprises20 and products is used for the price survey, consistent with the guidelines of the Statistics Committee of the CIS. Most important enterprises by class of activity (representing 60–70 percent of the class of activity) were selected first, followed by selection of the most representative products within the class. Small enterprises and household unincorporated enterprises are not included in the price survey, partly because of the threshold applied to select enterprises and partly because of their unstable character. As in other transition economies, the high mobility of the small and unincorporated enterprises makes it difficult to maintain a constant list of products and outlets throughout the year, and the Goskomstat prefers to exclude these from sampling. For a good representativeness of activities covering all production costs, a sample of small enterprises and unincorporated enterprises should be included in the PPI scope.

Products selection for the PPI basket is guided by their representativeness within reasonably homogenous groups in terms of key quality characteristics, individual properties, and availability over a long period. The range of products in the basket is rather limited (over 100). These products are priced monthly from 100 enterprises out of 400 in the state register, representing about 75–80 percent in total production. Different forms are used for the survey according to the activity. Price collection is done by mail.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

Source data for both weights and prices are not consistent with the time of recording, scope, and output valuation (see also 2.1.1.). In terms of classification, both PPI and the industrial output use the same classification of activities.

3.1.3 Source data are timely

Source data are timely. Both collection and processing of data are timely. The reporting forms should be submitted no later than the 3rd of the following month. The processing is done over the following week. The PPI production and dissemination is tightly correlated to that of the CPI and other indicators since they are published in the same publication.

3.2 Assessment of source data

3.2.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide statistical processes

The weighting structure is based on the production of all reporting industrial units, for which the response rate is close to 100 percent. The enterprises are compelled to submit timely data, a principle that is largely followed in practice. The producer price collection uses purposive sampling, so no sampling errors are available. Response rate is 100 percent; thus no adjustment is needed.

Monthly price quotes are checked with related products and for time consistency. The price survey form includes a column for comments and explanations. Unusual values for which there are no valid explanations are confirmed with respondents, who have to respond in writing. For important products, such as aluminum, the monthly price forms are signed by the director of the enterprise. Efforts are made, as possible, to maintain a representative sample of products and industries, given the limited size of the sample.

3.3 Statistical techniques

3.3.1 Data compilation employs sound statistical techniques to deal with data sources

Staff make efforts to ensure a high quality of collected and processed data to the extent possible. However, an accuracy check is mostly done manually before and after the upload of individual data because of the unsophisticated software available, and it depends greatly on the experience and judgment of staff. The monthly questionnaire requires that the respondent enterprise enter both the previous and the current month’s price. These data are compared to the previous month’s reporting form to help ensure that the same product is being priced. Respondents usually comment on the questionnaire when unusual price changes occur.

Temporary missing prices are handled by using any of the following methods: (1) impute the average price recorded during the past six months (mostly used); (2) use the movement of the next level of aggregation in the index, or (3) carry forward the last reported price (when prices are reported with some delays). No seasonal products are included in the PPI.

For disappearing products, replacements are introduced at the beginning of the next year when the weights are updated. The treatment applied differs depending on the point in time when the disappearance occurs. When products disappear at the beginning of the year, they are usually omitted from calculation, and where they have weights, these are redistributed within the group. If the products disappear sometime in the middle of the year, an average price of the covered period of the year is imputed. Both methods are liable to introduce bias, and preferably a replacement should be sought, for which quality adjustment is made. Prices for the replacement products are collected separately until the beginning of the next year when the replacement is introduced. The replacement product specifications are discussed with specialists who also give an estimate of the quality change to measure the real price increase. Examples of cases of replacement done by the Goskomstat refer mostly to the clothing industry (socks) and to household equipment (improved model of refrigerator). Reportedly, new products that do not replace existing products in the basket are not frequently introduced.

Most imputations appear to be done by the reporting enterprises. This leaves Goskomstat without the possibility to check the accuracy of the imputation and the time of recording, particularly in the cases where the reported price does not raise doubts. Concerted actions need to be taken by Goskomstat to instruct the reporting enterprises with regard to the price to be reported. All adjustments should be done by qualified staff at Goskomstat. Furthermore, the reporting forms for the industrial production used in the PPI compilation need to be improved (see also 2.1.1).

3.3.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

Annual industrial production is used as the basis for weighting the PPI, with deviations from the 1993 SNA gross output concept as noted in 2.1.1.

Elementary aggregates and higher aggregate indices are compiled using similar aggregation techniques as for the CPI. The only difference is that the PPI weights are annual production values of the previous (but one) year. The index aggregation at the higher level follows the following sequence: product–subindustry–industry–total index.

A conceptual discrepancy between the production coverage (domestic production plus exports) and the price collected (for domestic market) has been maintained over the past years. Conceptually, if the PPI covers exports, then export prices should be proportionally represented in the basket. In the case of Tajikistan, the exports should remain in the PPI because of their importance. The use of a domestic price for both export products and those for the domestic market has the potential of introducing serious bias in the index if the monthly evolution and trend of the two prices are different. Such is the case with aluminum, for which most of the production is exported (over 80 percent), and whose share in the PPI exceeds 50 percent. The exercise conducted by the mission on the two different prices21 demonstrated the urgency of correcting the formula used. A decision should be taken to introduce separate weights for the compilation of the monthly average price for aluminum. Retroactive calculation of the index series is indicated for as long as possible.

A similar treatment should be applied to the other products in the basket for which a part of the production is exported. The distribution of the annual production by destination will serve as a basis to derive such weights within the elementary aggregate. The same formula should be applied backwards to revise all prices used as denominators.

3.4 Assessment and validation of intermediate data and statistical outputs

3.4.1 Intermediate results are validated against other information where applicable

Monthly PPI series are compared for consistency with commensurate data from the CPI, import/export price indices, and other relevant price data.

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated

Unusual or unexpected index movements are investigated to determine their causes.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

There are no internal inconsistencies in the index because the imputations are done once only, and there is a single total for an industry made up of its relevant products.

3.5 Revision studies

3.5.1 Studies and analyses of revisions are carried out routinely and used internally to inform statistical processes (see also 4.3.3)

The PPI is final when first released, and the index is not subsequently revised. When methodology is revised, the index series are not revised backwards. A retroactive calculation should be undertaken to correct the problems noted above, and the results published, along with explanations and biases identified.

A regular analysis of the weights is done annually to assess eventual changes in the industrial production pattern. The output data used to determine the PPI weights are also compared with corresponding data from other sources.

4. Serviceability

4.1 Periodicity and timeliness

4.1.1 Periodicity follows dissemination standards

The PPI is compiled and published monthly, thus meeting GDDS standards.

4.1.2 Timeliness follows dissemination standards

The PPI is published no later than the 12th of the following month, thus exceeding the GDDS standards.

4.2 Consistency

4.2.1 Statistics are consistent within the dataset

The index is internally consistent since there is a single aggregation at the national level.

4.2.2 Statistics are consistent or reconcilable over a reasonable period of time

Consistent PPI series are available monthly back to 1994. Unusual changes in economic trends are explained in the analytical text included in publications.

4.2.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

The PPI is consistent with the other price indices, such as CPI, construction price index, import/export price index, etc.

4.3 Revision policy and practice

4.3.1 Revisions follow a regular and transparent schedule

The PPI weights are revised regularly (annually for the last years). This revision schedule is generally known by the main users.

4.3.2 Preliminary and/or revised data are clearly identified

The index number is considered final when first published. The publication does not mention the weights used for the PPI or their frequency of revision.

4.3.3 Studies and analyses of revisions are made public (see also 3.5.1)

The results of the revision studies, such as weights change, for the PPI are not published.

5. Accessibility

5.1 Data accessibility

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

The common price statistics publication contains clear tables and charts, as well as detailed analyses of PPI data. A reasonable amount of detail is published. The index is presented as the change from the last month and from the same month of the previous year. It also appears as a quarterly or yearly cumulated change.

5.1.2 Dissemination media and format are adequate

In addition to the monthly publication in the bulletin of social indicators, PPI series are published in a three-year publication on prices, as well as in the yearly book. It should be noted that the cost of publications is discouraging for the public at large. Moreover, requests from individuals are monitored, along with the subscribers, and these are questioned about their reasons to buy these publications. No Internet dissemination is available yet.

5.1.3 Statistics are released on a preannounced schedule

The PPI is disseminated according to a preannounced monthly calendar that can be found in the special catalog of information. This catalog presents the periodicity and the scheduled timeliness for Goskomstat’s publications.

5.1.4 Statistics are made available to all users at the same time

In principle, the PPI is released to all users simultaneously.

5.1.5 Statistics not routinely disseminated are made available upon request

Nonconfidential data at a higher level of detail are available upon request and for a fee. However, this fact is not widely known, and a preferential treatment seems to be applied for main users.

5.2 Metadata accessibility

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from internationally accepted standards, guidelines, or good practices are annotated

The metadata available in the monthly publication consist of a very brief description of the PPI. More details are provided in the three-year publication on price statistics but without any information about shortcomings in the data.

5.2.2 Levels of detail are adapted to the needs of the intended audience

Although a short documentation is available in various publications, a comprehensive document describing the concepts, sources and methods used for the collection and compilation of the PPI is not available. In order to better document details for internal and external users, it is recommended to produce and publish such a methodology. This methodology should be updated when methodological changes occur.

5.3 Assistance to users

5.3.1 Contact points for each subject field are publicized

All statistical products of the Goskomstat include a contact person, telephone number, mailing address, and e-mail address of the price unit. A prompt service is provided to inquirers.

5.3.2 Catalogs of publications, documents, and other services, including information on any changes, are widely available

A catalog of publications updated annually is available for consultation but not widely distributed. The main users, who are also subscribers, receive a copy of this catalog annually.

Table 3.

Republic of Tajikistan: Data Quality Assessment Framework (July 2003): Summary of Results for Producer Price Index

(Compiling Agency: State Committee on Statistics)

article image
article image

IV. Government Finance Statistics

0. Prerequisites of Quality

0.1 Legal and institutional environment

0.1.1 The responsibility for collecting, processing, and disseminating the statistics is clearly specified.

Government finance statistics (GFS) are compiled and disseminated under the general provisions of Public Finance Law No. 684 (June 28, 2002), and the state budget laws. These laws are supported by Cabinet Resolution No. 46 (February 5, 2003) that sets out the by laws governing the operations of the MOF. Among other things the by laws require the MOF to compile and disseminate fiscal data and give the MOF the right to demand the provision of data relating to public finance from other agencies.

The MOF supplies GFS to the State Committee on Statistics (Goskomstat) for publication. Goskomstat is free to determine the format and degree of detail in its publications but is not permitted to change the data. MOF also disseminates GFS expenditure data relating to health and education directly through the press, provides detailed GFS on request to individual users, and plans to include GFS on its website, which is currently under development.

The MOF, the National Bank of Tajikistan (NBT), and the Goskomstat agree that the MOF has primary responsibility for all fiscal data.

0.1.2 Data-sharing and coordination among data-producing agencies are adequate.

Under provisions of the law on the NBT, supported by a resolution pertaining to NBT’s Operational Department (OPERU), OPERU provides the Treasury Department of the MOF, on a daily basis, with data on receipts and payments to the Treasury single account, together with supporting documentation. This provides the basic data source used for the republic (central) and local government budget execution systems.

The MOF and NBT regularly meet to reconcile statistical data.

The central and local government budget systems cover all general government units, except for the Social Protection Fund (SPF), which has its own account in the National Savings Bank. Data required from the SPF to compile consolidated general government statistics are supplied to MOF monthly, as required under the provisions of the public budget laws, and the MOF by laws.

The MOF supplies data to the Goskomstat for publication and for use in compiling the national accounts, under the general provisions of the Law on State Statistics, which requires all ministries to provide data relating to the national economy to the Goskomstat.

Provisions for data-sharing among all agencies work satisfactorily.

0.1.3 Individual reporters’ data are to be kept confidential and used for statistical purposes only.

Tajikistan GFS covers only general government. The requirement to keep reporters’ data confidential does not apply to general government units.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response.

All fiscal data are obtained as a by-product of budget systems. No measures are required to encourage response to statistical surveys.

0.2 Resources

0.2.1 Staff, facilities, computing resources, and financing are commensurate with statistical programs.

There are no staff dedicated to GFS compilation or dissemination; the staff engaged in producing GFS have a wide range of other duties. Moreover, the process of producing GFS is dispersed throughout a number of MOF departments (principally the Treasury and Budget departments), and part of the basic compilation work (i.e., the provision to Treasury department of transaction data and supporting documentation) is carried out in OPERU.

The number of staff in MOF is adequate to carry out the compilation and dissemination of GFS at the level currently produced. A recent personnel review found that staff have sufficient skills to carry out their tasks (including production of GFS). However, OPERU staff resources are currently tightly stretched.

A general problem is salary levels throughout the whole public service, which makes it difficult to attract and retain the desired quality of staff and leads to undesirably rapid staff turnover. However, this is a problem that is not specific to MOF or OPERU.

The computer resources are adequate for the central (republican) budget level, with more than 80 percent of staff having access to a computer. However, computer resources are less satisfactory at the local budget level, with only between 30 percent and 40 percent of staff having computers. The extension of computer automation is a priority within MOF, and this situation is expected to improve in the near future. The transfer of data between OPERU and MOF is currently made using hard copy, because the software to allow electronic transfer has not yet been developed.

Physical facilities are basic but are adequate for the task.

Since the statistical function is not separated from other functions carried out by MOF, it is difficult to assess whether funding for the statistical program is reasonably secure. Funding is adequate at present, but because the statistical program is not a core task of MOF, there must be a risk that the statistical program would suffer if resources were severely constrained. While the MOF is required under its by laws to produce information on public finance, the extent, and degree of dissemination, of that information is not specified.

While the tasks carried out by OPERU do not appear to be those normally carried out by a central bank, they are essential to the current operation of the budget system and would be expected to be protected against resource cuts.

0.2.2 Measures to ensure efficient use of resources are implemented.

Staff performance reviews are carried out every three years (with the latest being November 2003) or more frequently if circumstances require it. Because statistical functions are not distinguished from other tasks, no special arrangements exist to improve statistical processes, although of course general procedures (such as computer automation) to improve staff productivity also flow through to statistical processing.

0.3 Relevance

0.3.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored.

GFS are widely used within the government as part of budget policy formulation and review, and outside of government to assess the governments’ fiscal policy stance and priorities. So far, no complaints have been received by MOF about the statistics provided.

However, the MOF does not actively seek out feedback from the users of GFS, nor does it have any mechanism in place to identify new or emerging data requirements. The Goskomstat sees its role as being confined to the publication of data supplied by the MOF.

The user survey conducted as part of the ROSC mission indicated that a majority of users (67 percent) considered that GFS were good (43 percent), very good (11 percent), or excellent (13 percent); while a substantial minority (33 percent) considered that GFS were unsatisfactory (22 percent) or poor (11 percent).

0.4 Other quality management

0.4.1 Processes are in place to focus on quality.

Statistical production is an integral part of MOF operations. GFS compilers follow established procedures and apply the same diligence as they do to other tasks. However, they lack a user perspective on the outputs of this process, and therefore the quality focus is on procedures rather than products.

0.4.2 Processes are in place to monitor the quality of the statistical program.

There are no processes in place to specifically monitor the quality of GFS, although MOF is open to feedback from users on the quality of the statistics. Review of the data would only occur as a result of specific events, such as indications by important users that significant problems existed or changes to the environment in which the statistics are compiled and disseminated.

0.4.3 Processes are in place to deal with quality considerations in planning the statistical program.

There are no specific processes in place to deal with quality considerations or statistical planning. Trade-offs in quality between timeliness and accuracy are implicitly taken into account in disseminating preliminary statistics, followed by statistics based on audited data. Because of the procedural focus referred to above, compilers are unaware of new and emerging data requirements.

The NBT accounting data, which is the primary source of GFS, is subject to annual international audit.

1. Integrity

1.1 Professionalism

1.1.1 Statistics are produced on an impartial basis.

There are no laws or other formal arrangements to support the professional independence of GFS compilers. However, in recent periods there has been no attempt by authorities to influence statistical outputs, and the government and other agencies appreciate the importance of allowing statistics to be produced free of interference.

The MOF does not generally recruit statistical specialists. However, because most staff have economic qualifications, in which a basic knowledge of statistical principles forms part of a wider set of skills, they do have a generally appropriate background for the tasks of compiling and using GFS. In addition, staff receive on-the-job training in carrying out statistical (as well as other) tasks. However, there is no formal training (i. e., using GFS experts) of MOF staff in GFS methodology and compilation methods. The MOF staff’s statistical qualifications are not sufficient for them to go beyond basic statistical compilation processes.

1.1.2 Choices of sources and statistical techniques as well as decisions about dissemination are informed solely by statistical considerations.

There is no evidence of political interference in the choices of the sources and statistical techniques in the compilation of the GFS. The administrative sources used in compiling GFS are obvious and appropriate, and the content and format of the statistics are based on IMF standards. Dissemination is done monthly in the Goskomstat’s bulletin Socio-Economic Situation in the Republic of Tajikistan, which is the standard publication for economic and social statistics for Tajikistan. In addition, selected expenditure statistics are published quarterly in the press.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics.

The MOF can, and occasionally does, respond to correct the misinterpretation or misuse of GFS.

1.2 Transparency

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public.

The terms and conditions under which GFS are compiled and disseminated are embedded in various laws and by laws and the internal administrative rules of the MOF and NBT. No information (apart from the laws themselves) on the terms and conditions applying to GFS is made public.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified.

The government has internal access to statistics prior to their release to the public but no information concerning this access is made public.

1.2.3 Products of statistical agencies/units are clearly identified as such.

The MOF is clearly identified as the source of GFS in Goskomstat publications and in press releases.

1.2.4 Advanced notice is given of major changes in methodology, source data, and statistical techniques.

GFS methodology, source data, and statistical techniques have not changed since 1999, when the current system was introduced. Advanced notice will be given of any future changes.

1.3 Ethical standards

1.3.1 Guidelines for staff behavior are in place and are well known to the staff.

There is no formal code-of-conduct for MOF staff. However, any such code would merely formalize the current practice expected from all staff. In practice there have not been significant problems with unethical behavior of MOF staff.

2. Methodological soundness

2.1 Concepts and definitions

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

Compilation of fiscal statistics for Tajikistan is based on the recommendations of A Manual on Government Finance Statistics 1986 (GFSM 1986).

There are no plans at present to migrate the basis of compilation to the new GFS standards based on the Government Finance Statistics Manual 2001 (GFSM 2001).

2.2 Scope

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices.

The general government sector in Tajikistan comprises the republican (central) budget, local budgets, and a social security fund (the Social Protection Fund—SPF). However, since the heads of the local governments are appointed by the central government, the local governments should be considered to be administrative entities controlled by the central government, rather than a separate level of government.

A number of relatively small “special funds” exist, which are controlled by budget agencies and are subaccounts of the single treasury account but are not included in the budget process. While aggregate data are available for these funds, at present no economic or functional detail is shown in GFS.

Statistics are produced for the republican budget. They are also produced for “State Government,” which is the consolidation of the republican and local budget sectors, and the SPF. State government therefore is equivalent to the general government sector for Tajikistan. No significant quasi-fiscal activity is carried out by public corporations. Therefore, the scope of fiscal statistics is consistent with the guidelines outlined in GFSM 1986.

Tajikistan fiscal statistics cover all the economic flows of general government to the standard specified in GFSM 1986. Statistics are also available on the domestic and foreign debt of the general government sector, in accordance with GFSM 1986.

2.3 Classification/sectorization

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices.

Although institutional sectors are not explicitly defined in Tajikistan’s fiscal statistics, the implicit general government sector (i.e., “State Government”) is consistent with general government in the 1993 SNA.

Statistics are prepared separately for central (republican) and local governments. As noted above, local governments are only an administrative segmentation of general government, and therefore the local government budget sector should be classified as an administrative subsector, not as a separate level of government.

Revenue, expenditure, lending minus repayments, and financing transactions are classified using the methodology set out in GFSM 1986.

2.4 Basis for recording

2.4.1 Prices used to value flows and stocks reflect actual or expected cash payments

All flow data are recorded on a cash basis, valued on the basis of the amount of payments and receipts in cash, as recommended in GFSM 1986.

The value of government debt is recorded as the amount borrowed, plus accrued interest. Although this is not in accordance with GFSM 1986, which recommends that debt be valued at face value (i.e., the amount due to be repaid at maturity), it is broadly in accordance with GFSM 2001, which recommends that all assets and liabilities be valued at market value.

Amounts denominated in foreign currency are converted to Tajikistan somoni using the official exchange rate for the end of each month, as provided by the NBT.

2.4.2 Recording is done on a cash basis

Economic flows are recorded on a cash basis, in accordance with GFSM 1986. All receipts and payments (covering revenue, expenditure, lending minus repayments, and financing) are recorded when reported as being credited (revenue, borrowing) or debited (expenditure, lending, amortization) to the single treasury account in the central bank, or, in the case of the SPF, its account in the National Savings Bank. Changes in the single treasury account in the central bank are reported daily.

There are no plans to move to an accrual basis of recording.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices.

Transactions are shown gross, except for corrective transactions such as tax refunds, which are recorded as negative revenue transactions, as recommended in GFSM 1986.

3. Accuracy and Reliability

3.1 Source data

3.1.1 Source data are obtained from comprehensive data collection programs that take into account country-specific conditions.

All transaction data for the republican and local government budget sectors are obtained from the budget execution systems. These systems provide complete and timely source data.

Complete and timely data for the SPF are obtained in hard copy form from the SPF accounting system, which has the same accounting structure as the budget systems.

The consolidation of the republican budget, local budgets, “special funds,” and SPF provides comprehensive coverage of the whole general government sector for Tajikistan.

Debt data are compiled from information in each individual loan contract, including the loan principal and schedule of payments. The foreign and domestic debt departments of the MOF control all government borrowing and monitor debt servicing transactions.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required.

The budget and SPF accounting systems are based on the economic and functional classification systems recommended in GFSM 1986. This allows GFS to be easily compiled from accounting data. The data are sufficiently detailed to allow GFS statistics to be produced at the classification levels recommended in GFSM 1986 and to allow consolidation of the various government administrative categories to produce general government statistics.

3.1.3 Source data are timely.

Source data for the budget sectors are very timely, being available generally within a few days after the transactions are recorded. Monthly data for the SPF are received within 10 days after the end of the previous month.

The complete, audited, final accounts of all government units are available within five months after the end of the reference year.

3.2 Assessment of source data

3.2.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and other nonsampling error; the results of the assessments are monitored and guide statistical processes.

Source data supplied by OPERU are checked against records of payment orders to ensure that the payments and receipts by the single treasury account are in accordance with approved payments by the Treasury Department of MOF. The data are also checked by the Ministry of State Revenues and Duties (MSRD). This process results in complete reconciliation of transactions and payment orders.

Transfers between different administrative sectors of government (e.g., between the republican and local budget sectors) are also checked and reconciled.

In addition, budget execution data are checked against budget planning estimates to identify any unexpected revenue or expenditure aggregates. This can be used to identify any major classification errors.

3.3 Statistical techniques

3.3.1 Data compilation employs sound statistical techniques to deal with data sources.

Preliminary data are routinely replaced by final annual data, and an appropriate balance is struck between timeliness and accuracy. The most timely data are sufficiently accurate and detailed for fiscal policy development.

Subannual (monthly and quarterly) data are benchmarked to final annual data.

3.3.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques.

There are no adjustments made to recorded data, which are already on the GFSM 1986 conceptual and classificatory basis.

3.4 Assessment and validation of intermediate data and statistical outputs

3.4.1 Intermediate results are validated against other information where applicable.

Not applicable to GFS.

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated.

Financing flow data on liabilities are not reconciled with changes in debt stock data.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated.

Discrepancies between fiscal data from the MOF and the MSRD, and monetary and financial statistics from the NBT, are checked and reconciled. The results of this exercise are made available to the Monetary Committee.

3.5 Revision studies

3.5.1 Studies and analyses of revisions are carried out routinely and used internally to inform statistical processes.

No studies or analyses of revisions are carried out. However, apart from the routine replacement of preliminary data by final annual data, no revisions have been made to GFS in recent times. Monthly and quarterly data are cumulative, and therefore no revisions to preliminary data are apparent. The changes resulting from replacement of preliminary by final annual data are usually small.

4. Serviceability

4.1 Periodicity and timeliness

4.1.1 Periodicity follows GDDS dissemination standards.

The periodicity of GFS meets, or exceeds, GDDS standards, except for debt statistics:

  • Central and general government transaction aggregates are disseminated monthly, quarterly, and annually (GDDS recommendation: quarterly and annual central government aggregates). However, the monthly and quarterly data are cumulative, rather than discrete.

  • Central and general government debt statistics are not disseminated in detail or on a regular basis.

4.1.2 Timeliness follows GDDS dissemination standards.

The timeliness of GFS meets GDDS standards, except for debt statistics:

  • Monthly and quarterly central and general government operations (revenue, expenditure, and lending minus repayments, and financing) are disseminated within two months after the end of the reference period. Cumulative monthly GFS are published in the Goskomstat’s monthly bulletin Socio-economic Situation of the Republic of Tajikistan, usually about six weeks after the reference period. (GDDS recommendation: quarterly central government aggregates within one quarter after the end of the reference period.)

  • Annual (cumulative monthly) central and general government data are disseminated within two months after the end of the reference period (GDDS recommendation: comprehensive annual central government data within six to nine months after the end of the reference period).

  • Central and general government debt statistics are not disseminated in detail or on a regular basis.

4.2 Consistency

4.2.1 Statistics are consistent within the dataset.

The same concepts and classifications are used for monthly, quarterly, and annual statistics. Subannual data add to annual data (for final annual, after benchmarking).

On an annual basis, the cash deficit/surplus is approximately equal to financing (with an opposite sign), and major aggregates are equal to the sum of their components. However, significant discrepancies exist between the cash deficit/surplus and financing in monthly data.

4.2.2 Statistics are consistent or reconcilable over a reasonable period of time.

A consistent time series is available for a period of five years. A major methodological change was introduced in 1999, and it is not possible to obtain consistent time series data before that date.

4.2.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks.

The financial sector transactions in GFS are consistent with the monetary and financial statistics.

GFS are used in compiling the general government sector of the national accounts without alteration.

4.3 Revision policy and practice

4.3.1 Revisions follow a regular and transparent schedule.

The preliminary-to-final revision cycle is predictable and stable, with preliminary data being replaced by final audited data on May 1 each year. Usually the revisions resulting from this practice are small and do not require explanation. The revision cycle would be known to all regular users of the statistics.

As only cumulative year-to-date preliminary data are published, revisions to previous months’ data are not visible to users. The internationally accepted practice is to publish discrete monthly data, in addition to cumulative year-to-date data, and to clearly identify revisions to historical data.

4.3.2 Preliminary and/or revised data are clearly identified.

Preliminary data are explicitly identified—all data not identified as preliminary are final.

To date, final data have not been revised.

4.3.3 Studies and analyses of revisions are made public.

No analyses of differences between the revised and preliminary data are produced because the revisions are routine and small.

5. Accessibility

5.1 Data accessibility

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts).

GFS are not disseminated according to GFSM 1986 recommendations. The GFS published by the Goskomstat are part of a compendium publication, and space constraints limit the amount of detail that is published. The main shortcomings of the published data are:

  • No summary tabulation showing aggregate revenue and grants, expenditure and lending minus repayments, deficit/surplus, and financing, is published.

  • A tabulation of expenditure and lending minus repayments classified by economic type is not provided.

  • Financing detail is not provided.

  • “Special funds” are not included in published statistics.

  • Monthly and quarterly data are cumulative.

  • Current monthly and quarterly data do not include time series for reference (time series are available for annual series).

Detailed statistics on government debt are provided only to the Parliament, government, and NBT. Total foreign debt is occasionally made public.

5.1.2 Dissemination media and format are adequate.

There is no dedicated GFS publication or website. GFS are disseminated only through Goskomstat hard-copy publications and MOF press releases with partial coverage of expenditure (health and education). Additional series, with greater levels of detail, are not published.

Goskomstat publications are available free of charge to official users but must be purchased by nonofficial users. The prices charged for publications are very substantial in relation to general income levels, and their availability does not appear to be widely publicized.

GFS are not currently available in electronic form. The MOF intends to include GFS on the MOF website, which is currently under development.

5.1.3 Statistics are released on a preannounced schedule.

Goskomstat publications are released on a regular and publicized schedule, but the timing of when the latest GFS data are included in the monthly bulletin can vary, depending on the date of release of data by MOF.

5.1.4 Statistics are made available to all users at the same time.

GFS are made available to official users before they are released to the public.

5.1.5 Statistics not routinely disseminated are made available upon request.

More comprehensive or detailed statistics than those published by the Goskomstat can only be obtained by application to the MOF. MOF is generally willing to provide more detailed GFS, but this obviously would depend on work pressure (and presumably also on the nature of the entity making the request). The availability of additional detail is not publicized.

5.2 Metadata accessibility

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from internationally accepted standards, guidelines, or good practices are annotated.

MOF has an internal compilers’ guide, and users can obtain information about the concepts and methodology of GFS on request to MOF. However, there is no published document that describes the concepts and methodology of GFS in Tajikistan.

A limited amount of information on concepts and methods is included in the GFS GDDS metadata, which is at present in draft form.

5.2.2 Levels of detail are adapted to the needs of the intended audience.

Only summary information on GFS concepts and methodology is available.

5.3 Assistance to users

5.3.1 Contact points for each subject field are publicized.

Liaison officers or contact points are included in the Goskomstat publication and MOF press releases. These contact details enable users to contact knowledgeable GFS staff.

5.3.2 Catalogs of publications, documents, and other services, including information on any charges, are widely available.

The Goskomstat has a catalog of publications and information on publication charges, but neither of these is widely available.

Table 4.

Republic of Tajikistan–Data Quality Assessment Framework: Summary Presentation of Results for Government Finance Statistics

(Compiling Agency: Ministry of Finance)

article image
article image
article image

V. Monetary Statistics

0. Prerequisites of quality

0.1 Legal and institutional environment

0.1.1 The responsibility for collecting, processing, and disseminating the statistics is clearly specified.

The National Bank of Tajikistan (NBT) is the sole official agency responsible for collecting, compiling, and disseminating monetary statistics. The authority of the NBT to request information from commercial banks and other credit institutions is granted by The Law on the National Bank of the Republic of Tajikistan (NBT Law) (amended as of December 14, 1996):

(1) Article 33—Banks and other credit institutions shall be obligated to provide the NBT with any requested information and data necessary for the discharge of its functions. The NBT shall have the right to publish this information and aggregate data in full or in part for different categories of banks and other credit institutions classified by the type of their activities;

(2) Article 37—The institutions having a permit to exchange currency, including banks, shall be obligated to submit to the NBT regular reports on their foreign exchange operations, including their foreign exchange exposure. The reporting procedures and the required documents shall be specified by the NBT; and

(3) Article 68—At the end of the fiscal year, the NBT shall draw up a report that includes a balance sheet, a profit-and-loss statement, a consolidated balance sheet for the banking system of the republic, and other necessary documents.

0.1.2 Data-sharing and coordination among data-producing agencies are adequate

Arrangements are well established among the NBT departments, as well as between the NBT and reporting institutions, to ensure the timely flow of source data for use in compilation of monetary statistics. The responsibility for compilation of monetary statistics resides with the Analytical Division of the NBT Monetary Policy and Statistics Department. Several other departments of the NBT support, directly or indirectly, the collection, compilation, and dissemination of monetary statistics. These are the (1) Accounting Department (aggregated balance sheet data of the NBT), (2) Bank Supervision Department (aggregated balance sheet data of commercial banks and nonbank credit institutions), (3) Foreign Exchange Department (international reserves), and (4) International Relations Department (foreign assets and foreign liabilities).

The balance sheet data of the NBT are produced by the Accounting Department and forwarded to the Analytical Division daily. That is, the end-month NBT balance sheet data will be available to the Analytical Division with a one-day timeliness. For the balance sheet data of commercial banks and nonbank credit institutions, the Bank Supervision Department receives individual monthly balance sheet data from commercial banks and nonbank credit institutions and prepares aggregated balance sheet data, which are available to the Analytical Division about 25 days after the end of the reference month.

As needed, the Analytical Division coordinates with the Bank Supervision Department to ensure that new data requirements are appropriately incorporated into the existing balance-sheet report forms for commercial banks and nonbank credit institutions.

0.1.3 Individual respondents’ data are to be kept confidential and used for statistical purposes only.

Monetary statistics are not disseminated on the basis of individual, primary data. By definition, monetary statistics refer to the secondary, analytical data that represent aggregated/consolidated, sectorized/classified information. To produce monetary statistics, the Analytical Division uses the IMF Statistics Department (STA) missions’ recommended computer program, which prevents disclosure of all individual balance sheet data.

At the time of accepting employment with the NBT, all staff are required to abide by the rules and regulations, with dismissal penalty, to prevent disclosure of confidential information to unauthorized users. The Law of the Republic of Tajikistan on Banks and Banking Activities also provides foundation to prevent disclosure of confidential information. To discharge responsibilities, the NBT employees are prohibited from disclosing or transferring to other persons information received in the course of their duty (Article 31.2). Also, Article 31.5 of this law states that all other ministries and agencies of the Republic of Tajikistan and their departments (divisions) cannot check banks’ activities.

The NBT has also taken several measures for preventing disclosure of individual balance sheet data, including (1) password-protected electronic data files that can only be accessed by staff involved in the data compilation; (2) password-protected computers to allow access to only authorized staff; and (3) security guards, who are posted at appropriate places to secure the NBT building.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

In the NBT Law, Article 33 states that banks and other credit institutions shall be obligated to provide the NBT with any requested information and data necessary for the discharge of its functions. Banks are required to submit their balance-sheet report forms to the Bank Supervision Department within 10 days after the end of the reference month.

Consistent with its legal authority, the NBT provides banks with written instructions for completing their monthly balance-sheet report forms. As needed, the Bank Supervision Department, in consultation with the Analytical Division, maintains regular dialogue with reporting banks to discuss, clarify, and resolve any data issues that may arise.

In the law of the Republic of Tajikistan on Banks and Banking Activities, Article 26.1 specifies that the NBT can cancel a license for carrying out banking operations in cases of (1) discovery of unauthentic reporting data and (2) carrying out banking transactions not envisaged by the license of the NBT. Also, in the case of noncompliance with required data provision, the NBT has established penalty that is rigorously enforced: $500 for misreporting and $100 per day for delayed reporting.

0.2 Resources

0.2.1 Staff, facilities, computing resources, and financing are commensurate with statistical programs

A core staff with adequate experience and training is maintained for the compilation of monetary statistics. The number and qualification of staff working on monetary statistics are adequate. The seven staff in the Analytical Division have primary responsibility for monetary statistics. They have college degrees in economics and related fields. In addition to on-the-job training, many have participated in the economics/statistics related courses conducted by the IMF’s Joint Vienna Institute (JVI).

The levels of salary of the NBT data compilation staff are adequate in comparison to the comparable workers in the commercial banks.

Not all staff in the Analytical Division are equipped with computers. There are four microcomputers, but only one is powerful enough to handle the compilation of monetary statistics. The whole process of data compilation and tabulation is computerized. However, there is inadequate provision of an emergency back-up system.

The Analytical Division is embarking on a new quarterly publication, Monetary Review, in addition to the NBT News. Also, the current NBT Bulletin will be augmented to include, among other things, a third language—English, in addition to Tajik and Russian. Thus, there is need for both new computer hardware and software to be commensurate with the increase in assigned tasks of the Analytical Division staff.

The office, furniture, and equipment, such as desks, chairs, filing cabinets, telephones, and related equipments need to be updated to reflect the number of staff and increased responsibilities. For example, currently, five staff of the Analytical Division are sharing four desks in an office.

Adequate financial resources are allocated to the Analytical Division for the work on monetary statistics. However, with the task of publishing the new Monetary Review in May 2004, financial resources may need to be reviewed to reflect this increased responsibility in the near future.

0.2.2 Measures to ensure efficient use of resources are implemented

The efficient use of existing resources is promoted by the Analytical Division through (1) annual performance review for staff; (2) training of staff, including courses at the JVI, in relevant areas to enhance technical knowledge and skills; (3) periodic review of work processes, as needed, to ensure consistent concept and classification for analytical accounts of the central bank and the commercial banks, and the monetary survey; (4) IMF technical assistance to assess statistical methodology and compilation system; and (5) allocation of financial resources toward prioritized tasks.

0.3 Relevance

0.3.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored

The Analytical Division undertakes several actions to address issues of concern to users. For example, on an ad hoc basis, it informally consults data users—such as the international organizations in Dushanbe, research institutes, and university students and professors—to monitor issues of usefulness of current classification, periodicity, and timeliness of monetary data, as well as to identify emerging requirements in this area. Also, NBT staff participate in statistical seminars sponsored by international organizations.

Although no formal arrangement for a user survey of monetary statistics has been undertaken, the results of the data ROSC User Survey in Tajikistan show that most respondents who use monetary statistics routinely are banks, universities and research institutes, enterprises, ministries and state agencies, and embassies.

0.4 Other quality management

0.4.1 Processes are in place to focus on quality.

The NBT management recognizes the importance of quality statistics as shown in their actions to (1) participate in the General Data Dissemination System (GDDS); (2) publish the new Monetary Review with occasional dissemination of methodology on compilation of data; and (3) train NBT staff on the importance of data quality and the method to achieve data quality.

Data are reviewed by the Monetary Policy Committee weekly to ensure quality. NBT management also reviews and discusses the development in monetary trends using the data compiled by the Analytical Division. Data are supplemented by charts to show trends over time. Any questions about the accuracy of the data from the management will be carefully reviewed and, if needed, errors corrected. Thus, the NBT management emphasizes the importance of quality with NBT managers held accountable for their performance.

The reliability and accuracy of balance sheet data of the NBT and banks—the primary source data for monetary statistics—are scrutinized through the external auditing process required by Article 69 of the NBT Law and Article 42 of The Law of the Republic of Tajikistan on Banks and Banking Activities. Article 69 of the NBT Law states that “Reports, accounts, and registers of the NBT may be subjected to an external audit carried out by authorized auditors.” For commercial banks, Article 42.3 states that an annual audit by an independent audit firm shall be carried out from the point of view of “timely, full, and exact reflection of banking transactions in accounting and reports.” The audit report is routinely published in the NBT Annual Report.

As needed, the Analytical Division, in collaboration with the Bank Supervision Department, periodically reviews, identifies, and resolves issues raised pertaining to the accuracy of monetary statistics. For example, assessments of banks’ reported data are routinely undertaken. Unusual data movement is reviewed for misclassification or error.

0.4.2 Processes are in place to monitor the quality of the statistical program.

The Analytical Division has established a monitoring process to identify and resolve issues raised at various stages of the compilation and dissemination of monetary statistics. It also periodically reviews the compilation process to identify any need for revision of data published in the NBT Bulletin. As needed, it provides corresponding documentation in footnotes relating to the revisions.

0.4.3 Processes are in place to deal with quality considerations in planning the statistical program.

As part of its work program, the Analytical Division routinely reviews and monitors accuracy of its monetary data.

Feedback from users, such as universities and research organizations, has been received routinely from Dushanbe and abroad. Many researchers have visited the Analytical Division to discuss data requirements for their research. NBT staff also conduct seminars periodically for commercial banks, discussing the need for accurate data and emerging requirements for monetary statistics.

NBT staff recognize the trade-offs among the dimensions of data quality. Increased attention has been given to improvement of timeliness and accuracy/reliability of monetary data in the context of preparation for the forthcoming participation in the GDDS.

1. Assurances of integrity

1.1 Professionalism

1.1.1 Statistics are compiled on an impartial basis

The professional independence of the NBT in carrying out statistical functions is authorized by Article 24 of the NBT Law, which states that “The National Bank of Tajikistan shall be independent in its activities within the limits of power granted to it by law.” Also, Article 6 of the NBT Law states that “The National Bank of Tajikistan shall organize and carry out its activities independently. No government interference in the organization of its activities shall be allowed.” Furthermore, Article 24 states that “Legislative and executive authorities shall not have the right to interfere in the activities of the National Bank of Tajikistan and its organizational units exercising the Bank’s legislatively confirmed powers. At the request of the National Bank of Tajikistan, any interference in its activities shall be examined by the Madjlisi Oli of the Republic of Tajikistan and the president of the Republic of Tajikistan.”

Article 24, ensuring that NBT management is supportive of the professional independence of NBT staff who compile and disseminate monetary statistics, states that “The National Bank of Tajikistan and the Government of the Republic of Tajikistan shall inform each other about the proposed actions of nationwide importance, coordinate their work, and hold regular consultations.”

NBT recruitment and promotion of staff are based on their relevant expertise in economics and statistics. Also, the NBT management provides on-the-job training in the monetary statistics methodology and compilation methods, including participation in seminars, courses, and workshops arranged by STA to further knowledge of statistical practices. In this connection, the STA’s technical assistance has played an important role to facilitate the training process. The NBT work process also promotes a culture of professionalism such as peer review of statistical work, publication of research, and analysis as evident in the forthcoming publication of the Monetary Review in May 2004.

1.1.2 Choices of sources and statistical techniques as well as decisions about dissemination are informed solely by statistical considerations

The choice of sources and statistical techniques for the compilation of monetary statistics is based solely on statistical considerations, i.e., balance sheet data of banks, as recommended by the IMF Monetary and Financial Statistics Manual (MFSM).

The NBT decisions to disseminate data are based solely on statistical considerations, without interference from other government agencies. The NBT Law allows, but does not require, the NBT to disseminate monetary data. Thus, the NBT disseminates monetary data as a service to the public.

The NBT decisions about the timing of data dissemination are based solely on statistical considerations, reflecting its internal advance release calendar, without interference from other government agencies. Routinely, the NBT publishes a wide range of monetary statistics with explanation and analysis to aid users in understanding the published data.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

When there is erroneous interpretation or misuse of the monetary statistics, the NBT staff would provide comments and clarification on a timely basis through the NBT publications, such as the monthly Bulletin, NBT News, and the new Monetary Review.

The NBT also releases data through news media, such as newspapers. Thus, it does monitor media coverage of monetary data.

1.2 Transparency

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

The laws and regulations under which statistics are collected, processed, and disseminated by the NBT are widely available to the public. The NBT Law, the Banks and Banking Activities Law, and other rules and regulations in the jurisdiction of the NBT are also available on the NBT Internet website, http://www.nbtj.org.

However, no single document or source on the terms and conditions applying to monetary statistics is made public, although selection of jurisdiction are available in various places and the underlying laws are publicly available.

Monetary statistics and policy analysis are featured in NBT publications, such as the monthly Bulletin, NBT News, and Monetary Review. Also, these publications identify the address of the NBT Internet website, where more information about the NBT can be found.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

For the monetary data, there is no internal government access by other governmental agencies prior to their release to the public.

In principle, the “access by other governmental agencies prior to data release” is only applicable to jointly-published data, where approval of all agencies is needed before data can be released to the public. In those cases, internal government access, which may occur on occasion, is not identified to users.

1.2.3 Products of statistical agencies/units are clearly identified as such

Monetary statistics are clearly identified as an NBT product, with its name and logo. When data produced by other agencies are published, the source is indicated in the footnotes. The NBT requests attribution when its statistics are used for reproduction or redissemination.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

The NBT does not give advance notice of major changes in methodology, source data, and statistical technique. It announces major changes concurrently with data dissemination.

1.3 Ethical standards

1.3.1 Guidelines for staff behavior are in place and are well known to the staff

Article 76 of the NBT Law provides the conditions for hiring and firing the employees of the NBT, their official obligations and rights, and disciplinary actions against them. These guidelines are published by the NBT in its Staff Rules and Regulations. Requirements on staff integrity, professional competence, and independence are well known to new staff, who are reminded periodically in various training and announcements to staff.

Article 77 of the NBT Law establishes that members of the Board and other employees of the NBT may not represent any business interests of a commercial, credit, or other nature if the latter conflict with their official obligations.

Article 80 of the NBT Law establishes that the NBT has an Internal Control Department (internal auditor), which shall (1) ensure observance of the rules of internal control by providing guidance to relevant NBT officials and, in case of departure from the rules, by reporting to the chairman of the NBT, who shall take the necessary action to correct any departure from such rules; and (2) exercise control over operations of the NBT for the purposes of ensuring strict compliance with legislation.

2. Methodological soundness

2.1 Concepts and definitions

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

In NBT, compilation of monetary statistics, key concepts, and definitions follow the IMF MFSM. The Analytical Division of the NBT Monetary Policy and Statistics Department has implemented the recommendations of STA’s “Report on Monetary and Financial Statistics Mission” (December 30, 2002).

Tajikistan’s monetary survey represents the consolidation of the analytical accounts of the NBT and of the commercial banks (including nonbank credit institutions).

The main aggregates of Tajikistan’s monetary survey are net foreign assets, net claims on general government, claims on the rest of economy, money, quasi-money, counterpart funds, and other items (net). As reflected in the International Financial Statistics, the compilation of claims on, and deposits of, government is based on the general government concept recommended by the “Report on Monetary and Financial Statistics Mission” (October 29–November 14, 2002, issued on December 30, 2002), as well as STA’s other earlier mission reports.

2.2 Scope

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices

The institutional coverage of the monetary statistics refers to the depository corporation survey (or the monetary survey) covering the NBT, all 14 domestic commercial banks, a branch of a foreign bank, and three nonbank credit institutions. The inclusion of the three nonbank credit institutions in the monetary survey reflects the recommendations of the STA’s mission report of December 30, 2002.

The depository corporation survey (or the monetary survey) represents the consolidated accounts of the NBT, resident commercial banks, a branch of foreign bank, and nonbank credit institutions. All resident commercial banks and nonbank credit institutions report their balance sheet data to the NBT for compilation of monetary statistics.

2.3 Classification/sectorization

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices

In Tajikistan’s monetary statistics, resident institutional units are grouped into sectors that have similar characteristics as specified in the MFSM, i.e., financial corporations (commercial banks, nonbank credit institutions), general government, and the rest of the economy (nonfinancial corporations, households, nonprofit institutions serving households). The “sectorization,” i.e., classification according to economic sector, also distinguishes between resident accounts and nonresident accounts. The Analytical Division staff understand the concept of “residency” criteria as described in the IMF’s Balance of Payments Manual, fifth edition (BPM5). The sectorization between resident accounts and nonresident accounts is reflected in banks’ report forms as well as the original computer program for compilation of monetary data provided by STA missions.

The classification scheme for financial instruments is based on the liquidity of financial instruments and the legal characteristics that describe the form of the underlying creditor/debtor relationship as specified in the MFSM, including monetary gold, and SDRs, currency and deposits, securities other than shares, loans, shares and other equity, insurance technical services, and other accounts receivable/payable. Currently, banks do not engage in financial derivatives transactions. As recommended by the MFSM, securities repurchase agreements (repos) are treated as collateralized loans.

In compilation of the monetary survey, the Analytical Division uses a computer program provided by STA’s missions. Over time, new accounts at more detailed levels have been created in the chart of accounts for the commercial banks. However, they were not appropriately reflected in the original computer program provided by the IMF missions, resulting in the unintended misclassification of nonresident accounts in the other assets and other liabilities of the analytical accounts of the commercial banks and, thus, in the other items (net) of Tajikistan’s monetary survey.

The Analytical Division staff are in the process of reviewing and updating, as needed, the original computer program for compilation of monetary statistics to appropriately incorporate and classify all new nonresident accounts in the foreign assets and foreign liabilities of the monetary survey.

2.4 Basis for recording

2.4.1 Market prices are used to value flows and stocks.

Tajikistan’s monetary statistics are consistent with the MFSM recommendations: (1) financial assets traded in the market on a regular basis are valued by using the price quotations directly from the markets. In Tajikistan, only interbank assets fall into this category as there are no other financial markets; (2) for securities, transactions are recorded on the basis of market prices, when new issues are sold and existing securities are redeemed; (3) for gold, the NBT balance-sheet data reflect the daily market price of gold; (4) loans are valued at book value without adjustment for expected losses; (5) valuation adjustments are classified in the valuation accounts in the balance sheets of the NBT and commercial banks, but not identified in the corresponding analytical accounts of the NBT and the analytical accounts of the commercial banks (including nonbank credit institutions); and (6) in the monthly NBT Statistical Bulletin, and the Monetary Review, footnotes are used to explain, as needed, compilation practices that diverged from MFSM recommendations.

Conversion of foreign-currency-denominated assets and liabilities into local currency is carried out using the balance-sheet date accounting exchange rates established by the NBT. For this purpose, the NBT and all banks use the end-period mid-point between buying and selling rates quoted by the NBT.

2.4.2 Recording is done on an accrual basis.

In Tajikistan’s monetary statistics, as recommended by the MFSM, (1) financial transactions are recorded on an accrual basis; (2) interest on financial assets or liabilities is treated as accruing continuously during the accounting period; (3) interest arrears, as applicable, are included in the value of the corresponding outstanding obligations; (4) arrears in the payments of fees/charges related to off-balance-sheet loans are included in accounts receivable/payable on the balance sheet of banks; and (5) financial transactions are recorded simultaneously between transactors.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices.

In Tajikistan’s monetary statistics, as recommended by the MFSM and expressed in the computer program for compilation of monetary statistics (1) as appropriate, some categories of data in the analytical accounts of the NBT (central bank survey) and the analytical accounts of the commercial banks including nonbank credit institutions (other depository corporations survey) are presented on a net basis, while other categories of data are shown on a gross basis; and (2) in the monetary survey (depository corporations survey), all interbank “claims on/liabilities to” are netted, while preserving the presentation of data on “claims on/liabilities to” other domestic sectors and nonresidents.

3. Accuracy and Reliability

3.1 Source data

3.1.1 Source data are obtained from comprehensive data collection programs that take into account country-specific conditions.

The data sources for compilation of monetary statistics are the balance-sheet data of the NBT and the balance-sheet data of all individual, resident commercial banks. The sources also include the nonbank credit institutions, because their activities also include deposit taking, i.e., their issuance of liabilities can be incorporated in the national definition of broad money. The inclusion of the nonbank credit institutions in monetary statistics is consistent with the recommendations of STA “Report on Monetary and Financial Statistics Mission” (October 29–November 14, 2002, issued on December 30, 2002).

The source data for resident commercial banks and nonbank credit institutions are the individual balance-sheet report forms submitted to the NBT Bank Supervision Department. The current chart of accounts for the commercial banks, being reviewed for new data requirements, provides sufficient detailed accounts for the compilation of monetary statistics in conformity with the MFSM recommendations.

Over time, new detailed accounts are created in the chart of accounts for the commercial banks, but they are not included in the original computer program for compilation of monetary statistics.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required.

The source data from the balance sheet of the NBT and the individual balance sheets of resident commercial banks and nonbank credit institutions provide a reasonable approximation to the definitions, scope, classification, sectorization, valuation, and time of recording required for monetary statistics.

As needed, the Analytical Division staff provide commercial banks with guidance for completing the balance-sheet report forms. They also maintain close contact with reporting banks to resolve data issues that have important implications for the accuracy of the accounting data, e.g., valuation methods and supplementary data sources. As needed, supplementary data on sales of new securities are used for verification of the corresponding balance sheet data.

3.1.3 Source data are timely

The balance sheet data of the NBT are produced by the Accounting Department and forwarded to the Analytical Division daily. That is, the end-month NBT balance-sheet data will be available to the Analytical Division with a one-day timeliness. For the balance-sheet data of commercial banks and nonbank credit institutions, the Bank Supervision Department receives individual monthly balance-sheet data from commercial banks and nonbank credit institutions and prepares aggregated balance-sheet data, which are available to the Analytical Division about 25 days after the end of the reference month.

Thus, the subsequently prepared analytical accounts of the NBT and the depository corporations survey (or the monetary survey) become available to meet the recommended standards of timeliness of the GDDS, i.e., one-to-two months for the analytical accounts of the NBT and one-to-three months for the depository corporations survey (or the monetary survey).

The timely reporting of data by the commercial banks and the nonbank credit institutions is rigorously enforced by the NBT, which has established a penalty of $100 per day for delayed reporting.

3.2 Assessment of source data

3.2.1 Source data—including censuses, sample surveys, and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide statistical processes.

The NBT Bank Supervision Department, in collaboration with the Analytical Division, has computerized procedures to facilitate monitoring of data accuracy in the balance sheets reported by individual commercial banks, including (1) cross-check and identification of out-of-trend movements of reported data; (2) verification whether the overall balance in the balance sheet is equal to the corresponding summary of individual accounts; (3) identification of negative asset items; (4) identification of possible misreporting against the corresponding NBT balance-sheet data; (5) confirmation of relatively large-value positions; and (6) analysis of previously reported data for possible revision, thereby updating the monetary data as needed.

3.3 Statistical techniques

3.3.1 Data compilation employs sound statistical techniques to deal with data sources.

The Analytical Division uses the computerized program prepared and recommended by STA’s mission of October/November 2002 for compilation of monetary statistics.

This program contains classification, sectorization, equations, main and detailed account numbers corresponding to the charts of accounts for the NBT and the commercial banks, and many other features that facilitate data entry, processing, and verification of aggregation/consolidation of data according to the MFSM recommendations. Tajikistan’s monetary statistics are based entirely on the data from the balance sheets of the NBT and the commercial banks and nonbank credit institutions. There is no adjustment for missing data. Thus, if needed, missing data will be explained in a footnote to the data.

3.3.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques.

The Analytical Division staff has implemented the recommendation of STA’s “Report on Monetary and Financial Statistics Mission” (October 29–November 14, 2002) to include data on nonbank credit institutions in the monetary statistics. Thus, Tajikistan’s monetary statistics are based entirely on the data from the balance sheets of the NBT and resident commercial banks and nonbank credit institutions. There is no adjustment for missing data. Thus, if needed, missing data will be explained in a footnote to the data.

Procedures are built into the balance sheets to check the accuracy of the conversion from the foreign-currency-denominated accounts to domestic-currency-denominated accounts.

There is no seasonal adjustments in Tajikistan’s monetary statistics.

3.4 Assessment and validation of intermediate data and statistical outputs

3.4.1 Intermediate data are validated against other information where applicable

As needed, the Analytical Division routinely reviews and verifies whether the government securities in the balance sheets of the NBT and commercial banks are as recorded in the securities database maintained at the NBT.

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated

The Analytical Division routinely assesses the (1) data on government securities in the balance sheets of the NBT and commercial banks against the securities database maintained at the NBT, and (2) data on interbank claims against corresponding interbank liabilities.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

The computer program for compilation of monetary statistics prepared and recommended by the STA mission of October/November 2002 contains procedures and formulas to investigate and verify errors in the classification/sectorization and discrepancies between the monetary statistics and the source data. Also, the Analytical Division has routinely reconciled reported financial flow data with changes in the corresponding stock data in the banks’ balance sheets.

3.5 Revision studies

3.5.1 Studies and analyses of revisions are carried out routinely and used internally to inform statistical processes (see also 4.3.3)

The Analytical Division has procedures in place to assess preliminary data against final data and, as needed, investigate the discrepancies between the two versions and make appropriate adjustments to the data. However, the Analytical Division has not undertaken the studies of (1) revision frequency and number of time series revised; (2) direction and magnitude of revisions; and (3) the sources of errors—all of which could be handled by its staff in the future.

4. Serviceability

4.1 Periodicity and timeliness

4.1.1 Periodicity follows dissemination standards

Data on the analytical accounts of the NBT (central bank survey) and the monetary survey (depository corporations survey—the analytical accounts of the NBT, commercial banks, and nonbank credit institutions) are disseminated with monthly periodicity, as recommended by the GDDS.

4.1.2 Timeliness follows dissemination standards

The monthly data on the analytical accounts of the NBT (central bank survey) and the monetary survey (depository corporations survey—the analytical accounts of the NBT, commercial banks and nonbank credit institutions) are disseminated within one month after the end of the reference month, as recommended by the GDDS.

4.2 Consistency

4.2.1 Statistics are consistent within the dataset

The Analytical Division has procedures in place to reconcile datasets: the NBT data on “credits/liabilities” can be reconciled with the corresponding aggregated balance-sheet data on “liabilities/credits” of commercial banks and nonbank credit institutions. However, flow of funds accounts have not been prepared for Tajikistan.

4.2.2 Statistics are consistent or reconcilable over a reasonable period of time

For Tajikistan’s monetary statistics, consistent time-series data are available from May 1995. Procedures are in place to reconstruct historical series for monetary statistics when changes in source data, methodology, or techniques are introduced.

Methodological footnotes that identify and explain the main breaks and discontinuities in the component time series for monetary statistics are routinely published in the monthly Statistical Bulletin and will be published in the Monetary Review. Also, unusual changes in economic/monetary trends are regularly explained in the commentary included in publications such as NBT News Bulletin and the Monetary Review.

4.2.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

On an ad hoc basis, the NBT undertakes reconciliation exercises, especially with balance of payments statistics. The changes in net foreign assets (i.e., change in gross-foreign assets less change in gross-foreign liabilities) in monetary statistics are largely reconcilable with the corresponding measure derivable from balance of payments statistics, i.e., monetary authorities’ assets—equity securities, debt securities, loans, currency and deposits, other assets, reserve assets less monetary authorities’ liabilities—debt securities, loans, currency and deposits, other liabilities. Data on international investment position are not yet compiled.

The central government’s records on the government deposits in and government borrowing from the depository corporations (the NBT, commercial banks, nonbank credit institutions) in government finance statistics are largely reconcilable with the comparable data in monetary statistics.

4.3 Revision policy and practice

4.3.1 Revisions follow a regular and transparent schedule

Currently, there is no revision policy. A revision policy is being studied by the Analytical Division for future implementation.

4.3.2 Preliminary and/or revised data are clearly identified

Users are informed whenever monetary data are preliminary or revised, in footnotes along with published data.

4.3.3 Studies and analyses of revisions are made public (see also 3.5.1)

When the monetary data are revised, there is explanation and analysis of differences between the revised and preliminary data, particularly in the new Monetary Review.

5. Accessibility

5.1 Data accessibility

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts.

The Analysis Division disseminates monetary statistics in a clear manner with charts and tables, with various details, and commentaries on current-period financial developments. However, there are no estimates for monetary statistics nor seasonally adjusted series.

5.1.2 Dissemination media and formats are adequate

The NBT disseminates monetary statistics through the Monetary Review, the Statistical Bulletin, and the NBT News Bulletin.

5.1.3 Statistics are released on a preannounced schedule

An internal advanced release calendar for monetary statistics is available for use by the Analytical Division and other NBT staff. However, there is no advance release calendar published for users of monetary statistics.

5.1.4 Statistics are made available to all users at the same time

Monetary statistics are made available to all interested users simultaneously through daily newspapers and the Weekly Businessman Newspaper.

5.1.5 Statistics not routinely disseminated are made available upon request

More detailed monetary statistics are not made available upon request by the public.

5.2 Metadata accessibility

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from internationally accepted standards, guidelines, or good practices are annotated

Comprehensive sources and methods document, including information on concepts, definitions, classifications, data sources, compilation methods, statistical techniques, and other relevant methodological aspects and procedures are published regularly in the NBT News Bulletin, the Monetary Review, and the Statistical Bulletin.

Metadata describing the monetary statistics will be prepared for participation in the GDDS.

5.2.2 Levels of detail are adapted to the needs of the intended audience

Currently, details are not adapted to the needs of users. Information is to be published in the Monetary Review.

5.3 Assistance to users

5.3.1 Contact points for each subject field is publicized

Practices are not observed currently but are to be implemented in the future.

5.3.2 Catalogs of publications, documents, and other services, including information on any charges, are widely available

Practices are not observed currently but are to be implemented in the future.

Table 5.

Tajikistan: Data Quality Assessment Framework (July 2003): Summary of Results for Monetary Statistics

(Compiling Agency: National Bank of Tajikistan)

article image
article image
article image

VI. Balance of Payments Statistics

0. Prerequisites of quality

0.1 Legal and institutional environment

0.1.1 The responsibility for collecting, processing, and disseminating the statistics is clearly specified

The legal environment for collecting and disseminating the balance of payments statistics is sound. The Law of the Republic of Tajikistan on the National Bank of Tajikistan (NBT) (amended as of December 14, 1996) assigns to the NBT the responsibility for compiling balance of payments statistics (Article 4.10) and endows it with a clear mandate to obtain information from commercial banks and other credit institutions (Article 33). In addition, the NBT is entitled to adopt Statutory Acts and Regulations that bind upon the government, executive authorities, banks and credit institutions, legal entities and individuals (Article 6).

The balance of payments statistics have been compiled by the Balance of Payments Statistics Unit (BOPSU) of the NBT since January 1997. The NBT is responsible for establishing the statistical methodology, reporting forms, data collection, verification, processing, compilation, and dissemination of balance of payments data. Working arrangements among government agencies that provide statistical data for balance of payments compilation are consistent with their assignments of responsibility. To contribute to the balance of payments statistics, the NBT has established smooth working arrangements with the Customs authority, the Goskomstat, the MOF, and other government agencies.

0.1.2 Data-sharing and coordination among data-producing agencies are adequate

Clear formal arrangements are in place to promote data-sharing and coordination among institutions that contribute to the production of balance of payments statistics. In May 2003, the Tajik authorities transformed the former Interagency Working Group on Coordinating Balance of Payments Compilation into a legal entity, the Interagency Commission for Coordinating Balance of Payments, Preparation and Compilation. The commission is chaired by the NBT and includes representatives of the Goskomstat, the MOF, the External Aid Coordination Center in the Office of the president, the Customs authority, the Ministry of Economy and Trade, and the Head of the Foreign Exchange Control Unit of the Office of the president of Tajikistan. The Commission is entitled to adopt executive decisions to carry out all relevant tasks in balance of payments statistics.

In addition, formal institutional arrangements, in the form of government decisions, organize the collection of data from private and public enterprises and from individual respondents. The main regulation is Decision 366 adopted by the government of Tajikistan On the Improved Compilation of the Balance of Payments Statistics of September 6, 2002. Decision 366 provides the main institutional arrangements that enable the NBT to receive data from different primary sources and to compile balance of payments statistics. Arrangements with Goskomstat derive from the Law on State Statistics, whereby providing that Goskomstat is the only statistical authority allowed to run surveys among enterprises and individual respondents. Under this arrangement, the Goskomstat uses a reporting form designed by the NBT to compile balance of payments data derived from enterprise reports. The Goskomstat runs a total of 12 surveys on behalf of the NBT that provide data used for balance of payments statistics. With other departments of the NBT, efficient coordination has been established to share information on official reserves and cross-checking of commercial banks’ account balances. Exchange of information with other agencies (e.g., MOF) is supported by common shared interests and routine contacts at the most technical level.

0.1.3 Individual reporters’ data are to be kept confidential and used for statistical purposes only

Absent from the Law on State Statistics and from the NBT Law is the important provision that collected information is to be used uniquely for statistics purposes. As a consequence, for data used in the compilation of balance of payments statistics, some data in principle may be used for other purposes, although this has not happened. However, the newly introduced enterprise form and the bank form mention that information is collected for statistical purposes only.

Procedures to prevent disclosure of individual reporters’ data are mostly based on the provisions of the Law on State Statistics, which does not sanction infractions of confidentiality, and those of the NBT Law, which offers only limited guarantees on individual data disclosure.

Under the provisions of the NBT Law, employees of the NBT have to report conflicts of interest in a number of cases (Article 76) and have to refrain from engaging in activities that may create a conflict of interest (Article 77), in which case no sanction is indicated. Information secrecy is enforced, except in the case when the persons’ expressed agreement is obtained to communicate the data (Article 78).

Neither the Law on State Statistics nor the NBT Law indicate any aggregation rules to prevent residual disclosure when aggregations of survey or other confidential data are disseminated; however, it seems that BOPSU staff is aware of special aggregation rules for information disseminated on the energy and aluminum sectors.

0.1.4 Statistical reporting is ensured through legal mandate and/or measures to encourage response

The legal framework organizes a shared authority over reporters of balance of payments statistics and clearly indicates that reporting is mandatory. The authority of the NBT to collect data from public entities, banks, and financial institutions is based on Articles 4.10 and 33 of the NBT Law, while the Law on State Statistics (Article 4) provides the Goskomstat with the authority to collect data from other public and private entities, which includes individual respondents and enterprises, irrespective of their form of ownership.

The Goskomstat collects all data pertaining to the activity of enterprises and individual respondents engaged in transactions with nonresidents through a total of 12 reporting forms and provides the NBT with aggregate information, in accordance with prescribed format and a delivery schedule set forth by a formal protocol between the NBT and the Goskomstat. The NBT uses this information to compile balance of payments statistics. Statistical information from banks, financial institutions, and administrative sources (chiefly the MOF and the Ministry of Economy and Trade) are directly collected by the NBT. If inconsistencies, ambiguities, or mistakes are found in statistical forms submitted by banks and financial institutions, BOPSU staff contact directly (by telephone or in writing) the institution to check the reported data or to offer guidance on proper transaction classification.

Compliance with data provision is ensured through strict penalties that however differ in their amount and enforcement, depending on the data collecting agency. Banks and financial institutions have to report monthly data to the NBT 20 days after the end of the reference month. Subsequently, to implement those provisions, the NBT has established a tariff for sanctions (for example, $500 in case of misreporting, $100 per day in case of delayed provision of information), which is rigorously enforced; the last penalties were imposed in 2003. The NBT deems that the sanctions have contributed to creating an excellent environment of cooperation from banks and financial institutions and to raising general awareness about its requirements in schedule and consistency for data delivery. For its part, the Goskomstat has adopted a proper system of sanctions. However, both the sanctions and their enforcement vary on a case by case basis. Compared to the NBT sanctions, those of Goskomstat are weaker in their amount (somoni 7 per infraction), and the scope of application is not uniformly enforced.

The way data collection is organized is conducive to creating a positive environment in goodwill and enhancing cooperation in data collection. For instance, the way of obtaining data from enterprises demonstrates consideration for respondents’ cooperation, since the Goskomstat is the only responsible institution in charge of collecting data from enterprises and individual respondents from the needs of both balance of payments and national accounts. This placement enables the Goskomstat to assess properly the reporting burden on respondents in terms of deadlines, scope, and coverage. Regarding collection of data from banks and financial institutions, cooperation builds up from (1) routine contacts at the technical level to resolve classification issues; (2) attendance at the annual training seminar on balance of payments statistics organized by the BOPSU; and (3) consultation when compilation methods or forms are introduced. Against this background, banks and financial institutions are expected to report both detailed statements to the Banking Supervision Unit of the NBT and balance of payments statements to the BOPSU, which eventually may imply duplication of efforts.

0.2 Resources

0.2.1 Staff, facilities, computing resources, and financing are commensurate with statistical programs

With nine professionals, staffing of the BOPSU is considered adequate to carry out existing tasks. Staff is recruited with at least a master’s degree in economics or information technology plus two-to-three years working experience. Working conditions, including salary, benefits, and status, have contributed to maintaining a challenging working environment for BOPSU staff while keeping a low turnover. Over the years, the Unit was able to build up solid capacity in compilation methods, expertise, and analysis in balance of payments statistics. BOPSU staff are lodged within the premises of the NBT and therefore benefit from adequate working facilities, including security of the premises. The NBT has received substantial technical assistance in balance of payments statistics in the framework of a TACIS22 initiative over the period 1997–2001 and from the IMF, with three missions during 2002–2003. This allowed continuous training of NBT staff as the country was engaging in privatization, modernization, and liberalization of its economy and financial markets. Funding is managed at the level of the NBT management and provides secure means to carry on identified needs for BOPSU’s statistical program on a medium-term horizon (two–three years).

In contrast, computers and other equipment are inadequate for the proper management and storage of data. Data are compiled in Excel and stored on the hard disks of the same computers without a back-up system and with the risk that files could be damaged during manipulation. The NBT management indicated that it plans to grant the BOPSU with new equipment.

0.2.2 Measures to ensure efficient use of resources are implemented

In principle, strategic planning in the NBT is subject to budget considerations and performance assessments; however, there is no internal process to measure overall resources used to compile statistics. For the purpose of collecting data from enterprises and individual respondents, the NBT has established synergies with other statistical agencies, for instance by sharing the task of collecting data on foreign direct investment through the wide network of local and central agencies of Goskomstat.

0.3 Relevance

0.3.1 The relevance and practical utility of existing statistics in meeting users’ needs are monitored

The NBT monitors closely the government’s needs in statistics; however there is no established consultation with a wider public of users. Historically, users’ needs have not been considered a priority, while compiling agencies are not mandated to consult a larger public of users. For instance, the NBT Law does not indicate that collection of statistics is intended for another purpose than economic policy, while the Law on State Statistics establishes a distinction between main users of statistics and a wider public (Article 9). Government and public agencies’ needs are monitored through the Interagency Commission for Coordinating Balance of Payments, Preparation and Compilation that gathers representatives from five agencies (see 0.1.2). In addition, the BOPSU conducts annual seminars on the methodological recommendations of the BPM5, which is attended by representatives of other agencies.

0.4 Other quality management

0.4.1 Processes are in place to focus on quality

Senior management of the NBT is sensitive to the quality of statistics. The BOPSU has acquired and implemented a software enabling an automatic processing of the monthly reports by banks and financial institutions and has built up its in-house team of specialists that maintain it. It has also established a recruitment policy for its staff, thereby ensuring a high standard of experience for the compilation of balance of payments statistics. Lastly, since it first started disseminating statistics, the BOPSU transposed the recommendations of BPM5 by taking into account the necessary transition to a market economy.

At the institutional level, the NBT management conducts a strategic planning meeting once a year, where decisions are taken on allocation of staff and funding. Besides, the chairman of the NBT has the function of advisor to the president of the Republic of Tajikistan. Regarding specifically management of accounts, the 2003 audit of the NBT found that bookkeeping procedures for accounts of banks, financial institutions, and other government entities, held on the books of the NBT, are soundly based on timely and quality control. Balances on these accounts are used to compile part of the financial statements used in balance of payments statistics for general government, commercial banks, and NBT operations.

0.4.2 Processes are in place to monitor the quality of the statistical program

The BOPSU has implemented several exercises to monitor and review the main features of balance of payments statistics. Surveys conducted with the support of the Goskomstat once every two years enable further improvement of the forms used to source data. For instance, the survey on foreign direct investment was revised in 2004 to collect stock data and valuation information to allow the compilation of an international investment position for the Republic of Tajikistan. Also, exchanges and contacts with other central banks provide the occasion to assess current techniques and procedures. This is evidenced against the background of foreign technical assistance: following hands-on training of NBT staff by European central banks (Banque de France in 1999 and Bundesbank in 2001), the NBT upgraded its automated system to collect and process data from commercial banks.

0.4.3 Processes are in place to deal with quality considerations in planning the statistical program

Senior management of the NBT recognize that there is a trade-off between timeliness and accuracy of the data. However, given that most financial account data are compiled on a monthly basis and the balance of payments statements are disseminated on a quarterly and on annual basis, this trade-off doesn’t impose significant tensions on the production schedule nor on the scope of the available data at the processing step. There is also a recognition that when the collection system was established, the NBT had to balance a tension between involving users, assessing alternative ways of data collection, and adhering to timeliness in disseminating data. This resulted in a trade-off that gave priority to the requirements of a restricted category of users and gave prominence to mandated procedures over goodwill and cooperation among reporters.

1. Assurances of integrity

1.1 Professionalism

1.1.1 Statistics are produced on an impartial basis

Statistical regulations and policies support the production of statistics on an independent basis. The NBT Law stipulates that the NBT enjoys independence as a distinct legal entity (Article 1). The NBT is managed by a Board and a chairman; the Board is the supreme collective body determining the main direction of the NBT. The chairman is appointed and dismissed with the agreement of the House of Representatives (Madjilisi Oli) (Article 71). Against this background, the NBT is able to choose freely its working methods and procedures in choosing data sources, methodological techniques, and cooperation with other entities. These provisions are further detailed in the NBT staff Code of Conduct, provided to each staff upon entry on work. For its part, the Law on State Statistics indicates that the methodology is based on scientific basis and imparts objectivity, credibility and independence to statistical information and to actuality of collection, stability of the data system and integrity in the information system (Article 1). At the NBT as well as at the Goskomstat, professional competency plays an important role in staff promotion; staff are encouraged to present seminars or share information on the work they have done. All but two BOPSU staff have participated in training organized by other central banks or by the IMF.

1.1.2 Choices of sources and statistical techniques as well as decisions about dissemination are informed solely by statistical considerations

The choice of source data is based solely on statistical considerations. The various sources used by BOPSU for the compilation of balance of payments statistics allow cross-checking. For instance, data on external debt communicated by the MOF are confirmed by bank reports. The data sources are similar to those used in neighboring countries and mix banking, administrative, and survey-based sources.

The dissemination of quarterly balance of payments data follows a regular schedule prescribed by the government decision of June 26, 1998 (quarterly and annual, three months after the end of the period). The mode of dissemination solely obeys the requirements of users for the format and detail of statistics and follows a standard layout as recommended by the BPM5 for standard components and analytical presentation.

1.1.3 The appropriate statistical entity is entitled to comment on erroneous interpretation and misuse of statistics

Balance of payments compilers actively follow the financial and economic press. The NBT has not experienced misinterpretation or misuse of statistics; however, as in other instances, it has the authority to follow up with comments and corrective information.

1.2 Transparency

1.2.1 The terms and conditions under which statistics are collected, processed, and disseminated are available to the public

Few elements are provided to the public about the way statistics are collected, processed, and disseminated. Excerpts of relevant legal material (e.g., laws) are usually printed in the press, and integral versions of laws are published in their totality in the legal announcement publication Djumbuhuriyat. However, other legal information is seldom disseminated; confidentiality of individual data, banking, business and state secrecy are not addressed in the statements made available to the public whether on the NBT Internet website, publications, or other public document.

1.2.2 Internal governmental access to statistics prior to their release is publicly identified

No information is made public on who has access to data prior to the date of the public dissemination. Once data are finalized by the NBT, they are sent to the government and a few other public authorities before the date of the press release. This practice is not made public because the authorities assume that the public is aware of prior information of government officials.

1.2.3 Products of statistical agencies/units are clearly identified as such

Hard copies and electronic publications of balance of payments data are clearly identified as NBT, with its name and logo. The balance of payments data, tables, and graphics identify the NBT as the source. However, while the Law on State Statistics mandates prior agreement by the Goskomstat before statistical information is reproduced and mandates explicit indication of the source (Article 7), there is no such requirement from the NBT for the dissemination of balance of payments statistics.

1.2.4 Advance notice is given of major changes in methodology, source data, and statistical techniques

No advance notice is given to users on main changes affecting data. However, comments are inserted as changes in methodology, source data, and statistical techniques are disseminated for the first time, whether as further developments in the data analysis or in the form of footnotes to tables.

1.3 Ethical standards

1.3.1 Guidelines for staff behavior are in place and are well known to the staff

The NBT provides staff with commensurate information on the ethical standards of the NBT. The NBT Law addresses general tenure conditions (Article 76), conflict of interest (Article 77), and professional secrecy (Article 78). In addition, an internal Code of Conduct is enforced. Staff are made aware of the ethical guidelines upon recruitment and during their tenure, for instance, during the weekly meeting that is held for all staff of the NBT.

2. Methodological soundness

2.1 Concepts and definitions

2.1.1 The overall structure in terms of concepts and definitions follows internationally accepted standards, guidelines, or good practices

The overall structure of the balance of payments statistics broadly conforms with the guidelines presented in the BPM5. However, in a number of areas, the concepts and definitions of BPM5 are not currently followed:

Residency: The legal definition of residency in Tajikistan is not in accordance with the concept of residency adopted in the BPM5 and in the 1993 SNA. The Law on Currency Regulation (1997) stipulates that residents are considered to be all legal entities established in the territory of Tajikistan, including their branches and representation offices in and outside the Republic of Tajikistan. This definition includes the local branch of a nonresident enterprise or bank as nonresident, which creates confusion for respondents. In addition, the Law on Currency Regulation is implemented by banks and financial institutions, while enterprises follow the definition of residents as indicated in the instructions that the Goskomstat had elaborated with the NBT and which mandates reporting according to the definition of the BPM5. The coexistence of two distinct definitions of residency burdens the task of the balance of payments compilers since it imposes complex reclassifications and hampers potential quality control. Data reported directly by enterprises cannot be properly confirmed through bank settlements for important transactions, for instance in the case of income for foreign direct investment.

Loans: The instruction to the enterprises lists the definitions used to report data according to the standard components of BPM5. However, the definition used for the category of loans is misleading (“non marketed contractual agreements or allocations to finance exports and imports in goods and services, disbursed by entities that are distinct from creditors”) and hampers a proper recording of data.

2.2 Scope

2.2.1 The scope is broadly consistent with internationally accepted standards, guidelines, or good practices

The balance of payments statistics include all foreign economic operations of all economic sectors within the geographic territory of Tajikistan, in accordance with BPM5. The NBT uses the comprehensive registers of all operating banks and enterprises, regardless of their form of ownership and whether or not they conduct international transactions, to collect data for balance of payments purposes. In addition, since 2001, with the support of the Goskomstat to complete the scope of balance of payments data, it has set up several surveys such as foreign direct investment, construction services, procurement of goods in ports and airports, and travel services.

The following exceptions, however, limit the scope of data:

Goods: For shuttle trade, the NBT has implemented surveys to measure imports and exports, focusing on construction goods that are not recorded by customs authorities. However, few surveys are undertaken to assess travelers’ expenses. In addition, as imports of goods are based on reports by customs posts deprived of proper computers, they rely on outdated registers of prices that contribute to undermining the value of imported goods. Also, barter trade, a major component of Tajikistan’s external activities, is valued using domestic prices that may not capture the higher prices prevailing on international markets.

Compensation of employees: Absent a survey of expatriate workers, balance of payments statistics do not capture the amounts received by Tajik employees who commute seasonally between Tajikistan and neighboring countries.

Investment income: Income data from joint-venture direct investment enterprises are currently not reported for balance of payments compilation. In addition, very few of the responding enterprises report income, since the Goskomstat may reveal this information to the Ministry of Revenue.

Current transfers: The External Aid Coordination Center has the responsibility to monitor the activities of donor countries and international organizations providing humanitarian grants and technical and development assistance to Tajikistan. For lack of response to the 53 forms that the NBT sends out, current transfers are underestimated to assess general government’ transfers. For workers’ remittances, only transfers transiting by banks are being recorded while transfers using individuals’ bank accounts in same countries (known as Awala system) miss the proper identification of the operations. In the absence of a survey of migrants, substantial amounts would still be missing.

Government services: While the expenses of Tajikistan’s embassies and representations abroad are correctly reported by the MOF, the expenses of foreign embassies and representations in Tajikistan are captured through banking reports, which miss transactions paid directly through accounts held with nonresident banks and cash transactions. A survey of foreign embassies and representations would help to strengthen reported data.

Other investment: Trade credits are reported by the banks but are incomplete on external private sector loans to finance cotton production, since information is missing on financial arrangements in that sector. In fact, Tajik cotton producers can obtain adequate funding from nonresident banks on the basis of the next harvest and use for this purpose future contracts. However, these arrangements are not made known to resident banks, which impedes the scope of balance of payments data. A revised form is needed to collect the missing information.

2.3 Classification/sectorization

2.3.1 Classification/sectorization systems used are broadly consistent with internationally accepted standards, guidelines, or good practices

The classification/sectorization system follows the standard components presentation of the BPM5, as well as the institutional, financial instrument, and maturity for financial instruments. In addition, the NBT has introduced a number of improvements since 2001. For instance, in bank reports, the subsectors for joint-venture enterprises, public enterprises and nonfinancial enterprises are separately distinguished within the other sectors. Also, the reporting form incorporates the standard balance of payments classification in bank reports and, since 2004, in enterprise reports. However, a few other issues remain pending:

Public sector external debt: Debt of general government and debt of public sector enterprises are not classified distinctly, whereas the BPM5 requires a separate recording for both components as follows: guaranteed and nonguaranteed debt of public sector enterprises is classified under other sectors, while government debt is shown separately under general government. As a result, the balance of payments data underestimate the external debt for public sector enterprises and overestimate the total for general government. Also, this creates an inconsistency in sectoral cross-checking with government finance statistics on external debt.

Nonmonetary gold is being classified with goods and is not shown separately in the standard balance of payments presentation.

Workers’ remittances: After a survey of bank clients in February 2004, the NBT plans to revise the classification of workers’ remittances as follows: all transfers below $1,000 and half of those between $1,000 and $3,000 should be recorded as workers’ remittances, while the rest should be classified as exports of goods. This supports a substantial revision of data for workers’ remittances.

Derivatives are not shown separately from portfolio investment. However, their use is extremely limited.

Other investment: For loans, cotton crop financing received from abroad should be broken down into repayments of principal and interest. For currency and deposits of banks, the balances of nostro and vostro accounts of banks are reported without any breakdown by sector and maturity.

2.4 Basis for recording

2.4.1 Market prices are used to value flows and stocks

Balance of payments data are compiled in millions of U.S. dollars. For transactions conducted through bank, transactions denominated in national and other currencies are converted into U.S. dollars using the exchange rate prevailing at the time of the operation for all transactions in flows. For transactions derived from stocks valued on a quarterly (direct and portfolio investment) basis, transactions reported by enterprises and individual respondents are valued at the exchange rate prevailing at the end of the quarterly period. For transactions derived from monthly reports (other investment and international reserves), the average monthly or quarterly exchange rate is used.

Transactions are, whenever possible, valued at market prices at the time the legal ownership is transferred. However, certain goods (aluminum, gas, oil and lubricants, electricity, and cotton) and services (rail and pipeline transport) are valued at contract prices since they are sold over-the-counter or in barter trade. In addition, humanitarian aid in the form of grants in kind (e.g., delivery of food) is valued by the donors and allows a wide array of prices that are not statistically revalued as they enter the Tajik economic territory. These practices may explain discrepancies with observed international market prices.

Exports and imports of goods are valued on an f.o.b. basis.

In 2004, the NBT introduced a revised form to collect foreign direct investment. That form includes the reporting of all prevailing valuation on stocks and transactions.

2.4.2 Recording is done on an accrual basis

Recording observes accrual principles.

In accordance with the recommendations of BPM5, goods are recorded at the time of customs clearance, and most services are recorded when they are rendered in accordance with the recommendations of BPM5. For government and government-guaranteed loans, loan drawings are recorded at the time of their actual disbursement, while installments of interest and principal are recorded on an accrual basis since the NBT bases its records on tables of payment schedule and amortization in capital. Dividends on investment and interest due on portfolio assets are recorded on a cash basis when the bank reports the settlement. For all other loans (those that were not registered by the MOF), payments of interest and capital are recorded when the reporting banks communicate the information to the NBT.

2.4.3 Grossing/netting procedures are broadly consistent with internationally accepted standards, guidelines, or good practices

The grossing and netting procedures are broadly consistent with practices recommended by BPM5. Data are compiled on a gross basis and are shown in gross terms in the current account and in net terms in the capital and financial account. Assets and liabilities components are shown separately in the published statistics; however, they can be provided in gross figures.

3. Accuracy and reliability

3.1 Source data

3.1.1 Source data are obtained from comprehensive data collection programs that take into account country-specific conditions

The NBT employs a comprehensive framework of data collection that includes the International Transactions Reporting System (ITRS), which reports surveys of enterprises and individual respondents and administrative data sources within and outside the NBT.

The ITRS is the primary data collection system used for compiling balance of payments statistics. It is based on detailed monthly transactions reported following the standard components of the BPM5 both in national currency (the Somoni) and U.S. dollars. Introduced in 1999, the ITRS covers reporting from 13 operating commercial banks23 in application of (continued) the law on currency regulation. ITRS data are mostly used to record current account data in the categories of services, income, and nongovernmental transfers. In the financial account, ITRS data are used to compile portfolio and other investment, while surveys contribute for specific services, such as transportation, communication, travel, insurance, operational leasing, and part of government services.

The ITRS reports individual transactions for each reporting bank in gross terms on a standard form and uses electronic transmission (email and diskettes). It allows automatic processing with imbedded algorithms that control the consistency of information by country, resident, sector, transaction codes, and amounts in somoni and U.S. dollars per day of transaction. The transaction nomenclature of 177 codes, which covers the full range of BPM5 standard classification, was revised by an NBT Instruction of April 2001. Data control and processing is supported by software “ISO,” developed by a Russian private company and maintained by in-house NBT IT experts.

Goskomstat compiles a total of 10 reporting surveys by the in application of Decision 366 (see 0.1.1). The main survey is the Investment Survey, which is collected from 100 of the 236 registered foreign direct investment enterprises. In general, the survey framework is dependable with complete entries, such as reinvested earnings. The form on which data are collected was designed by the NBT. In 2004, the form and the accompanying instructions were both revised to allow for the collection of stock and valuation data to produce the international investment position. Corresponding data were collected for the first time during the first quarter 2004. Other surveys managed by the Goskomstat include transportations services, foreign visitors in Tajikistan, humanitarian aid, foreign currency circulation, communication services, construction services, joint-venture enterprises, enterprise debt, and migrant workers. In addition, the Goskomstat collects data from the customs administration and supplements them with other sources to provide the NBT with commensurate information to compile the goods items of the balance of payments.

All survey results are provided to the NBT in hard-copy format. Data classifications necessitate further treatment and manual handling of the data.

In addition, the NBT runs its own survey to calculate the correction factor for c.i.f./f.o.b.

Among the administrative data sources, the MOF contributes with data on external debt, position in IMF borrowings, and government income and expenses with nonresident entities. For their part, the other departments of the NBT provide information that permits the calculation of international reserves and foreign currency liquidity for the monetary authorities and government. The departments also provide balance statements in foreign currency for the NBT and all licensed commercial banks.

In addition to those data sources, the NBT monitors regularly the financial press, mostly to review the information publicly available on significant current account transactions. This also allows the NBT to use outstanding transactions to illustrate the analysis it makes publicly available.

3.1.2 Source data reasonably approximate the definitions, scope, classifications, valuation, and time of recording required

In general, source data reasonably approximate the definitions, scope, classification and time of recording. The NBT has developed the following specific procedures to adjust data from various sources:

Goods

Data for exports and imports are obtained from the customs administration on a quarterly report form developed by the NBT. Goskomstat adds cross-border flows of electricity and gas, and humanitarian aid. For its part, the NBT supplements an estimate of shuttle and border trade and makes corrections to prices (for instance because the prices of Tajik imports have lagged world markets, to correct underestimated prices in barter and over-the-counter trade). The NBT also uses data that Goskomstat collects regarding goods and supplies acquired abroad by Tajikistan Airlines and Tajikistan Railways, as well as repairs performed abroad. These are used to compile repairs on goods in the current account. On behalf of the NBT, the Goskomstat conducts a survey to assess the c.i.f./f.o.b. adjustment factor every three years to derive estimates of freight and insurance costs in imported goods.

Services

Transportation: the Goskomstat collects statistics services from various report forms submitted by government agencies, state enterprises, other official institutions, as well as private firms, regardless of their capital share between residents and nonresidents. Data on transportation, including freight, passenger services, and other transportation, are provided by the Ministry of Transport, which includes Tajikistan Railways, Tajikistan Airlines, and road transport enterprises.

Travel: Expenditures for official travel are derived from the budget and are provided by the MOF. They include the expenditures of officials traveling abroad. Personal travel is communicated by tourism companies and covers only hotel expenditures abroad. The credit entries for personal travel are compiled from reports by tourism agencies, hotels, and the government dacha, which caters exclusively to official travelers.

Education services: Banks supply information covering tuition fees of foreign students in Tajikistan and tuition fees for Tajik students studying abroad. These transactions are classified correctly under travel.

Communication: The Tajik Telecommunication Company provides data on international communications. This information is supplemented by reports submitted by banks.

Insurance: In addition to data derived from the f.o.b./c.i.f. adjustment to assess merchandise insurance, the Tajikistan Life Insurance Company supplies information on insurance sold to nonresidents.

Operational leasing: The Ministry of Transport provides operational leasing data for both Tajikistan Airlines and Tajikistan Railways.

Government services: The information is supplied by the MOF for all expenditures of Tajik embassies and representations abroad, including wages and salaries of local staff and operating expenditures, such as fuel, office supplies, utilities, and other outlays. Embassies and international organizations have failed to respond to questionnaires regarding their corresponding expenditures in Tajikistan. This information is, however, estimated by the NBT through reports submitted by banks.

Other services: Banks provide information on financial, construction, and computer services.

Income

Compensation of employees: Using information provided by banks, the NBI estimates credits including wage and salary payments by embassies and international organizations to Tajik staff. Debits are obtained from MOF reports, as noted above, supplemented by data from banks.

Investment income: Banks provide information to the NBT on their interest income receipts and payments. From the NBT accounts, BOPSU staff obtains data on interest earnings from reserve assets and on interest payments on loans to the NBT from foreign institutions. The MOF provides data on interest payments on government and state enterprises, external debt, as well as on publicly-guaranteed debt. Other debt (e.g., nonguaranteed by state or local governments) is captured through bank settlements. However, up until the revised survey on direct investment in 2004, no data were reported by direct investment enterprises on interest payments to parent firms or on income payment.

Current transfers: The counterpart of humanitarian aid imports is reflected by credit entries under current transfers. Information on current transfers made by the government of Tajikistan (debit) is provided by the MOF, while the banks supply data on private transfers and on remittances from Tajik workers employed abroad (credits).

Capital account

The MOF provides data on external debt forgiven.

Direct investment: The Goskomstat collects data on registered direct investment enterprises for inflows, including loans from parent enterprises abroad and reinvested earnings. Since 2004, the revised survey allows supplementary information on changes in valuation.

Other investment: Trade credits are reported by the banks on a separate form, which leaves out enterprises operating in cotton production. Loans to government agencies and state enterprises, as well as government-guaranteed loans, are reported by the MOF on a special form designed by the NBT. The remainder of other investment is provided by banks as balances on nostro and vostro accounts.

Reserve assets: The balance sheet of the NBT provides this information on a monthly basis. The information includes a valuation at end-of-month for the exchange rate and change in value for the relevant financial asset.

3.1.3 Source data are timely

The NBT has established collection procedures to ensure that reporters meet timeliness requirements for data. As indicated in the reporting form and accompanying instructions, banks have to submit their full reports 20 days after the end of the reference month, and any delay involves follow-up by phone and incurs strictly enforced sanctions. In addition, the reporting form refers to the code and warns respondents that their civil responsibility is involved in case of delayed or misleading information.24 Information from Goskomstat is provided within limits of delays prescribed by Decision 366 On the Improved Compilation of the Balance of Payments Statistics. Data on enterprises, where nonresidents have foreign direct and portfolio investment, are provided 35 days after the end of the reference month, and all other surveys run by the Goskomstat on behalf of the NBT are communicated 45 days after the end of the reference month.

3.2 Assessment of source data

3.2.1 Source data—including censuses, sample surveys and administrative records—are routinely assessed, e.g., for coverage, sample error, response error, and nonsampling error; the results of the assessments are monitored and made available to guide statistical processes

The NBT identifies any errors, inconsistencies, and large or unusual transactions and follows up with the primary respondents for bank reports, with the Goskomstat in the case of surveys, and with the public agency in the case of administrative data. The NBT identifies inconsistencies on bank reports through its automated system that compares new data to those of former periods and detects possible discrepancies as data are communicated, and confirms with banks the accuracy of the reported figures. Data reported by the Goskomstat are checked against similar reports in the former periods and with reports from companies in the same sector and are eventually discussed with the Goskomstat head-office in Dushanbe. Data from administrations reported directly to the NBT undergo similar checks and controls. When two sources of information exist, the NBT assesses both sources to validate the reported information. For instance, foreign direct investment income is provided by bank settlements and by survey results. Only checked information is merged in the BOPSU database for balance of payments statistics.

3.3 Statistical techniques

3.3.1 Data compilation employs sound statistical techniques to deal with data sources

The data compilation techniques used by the NBT give priority to automated processing for the assessment of bank reports, which limits the risk of coding, editing, and tabulation errors, and the need for adjustment in case of nonresponse. Reports for former periods are used to benchmark new reports. Goskomstat data are checked in close cooperation with the local authority responsible for supervising the primary respondent, whether a company or individual respondent. Given Goskomstat’s constraints in time and resources to check nonresponse or follow-up on reporting errors, priority is given to individual reports that bear a significant contribution to the subitem total. In practical terms, a Goskomstat mission will visit the respondent if at least 50 percent of a given subitem is missing; this considerably limits the practical follow-up to exceptional visits by a Goskomstat controller. Data reported from administrative sources to the NBT undergo systematic checks and comparison against similar reports of former periods. In that case, data are adjusted after they are reviewed with the respondent by phone or by writing.

The NBT encourages the banks to limit the use of uncoded transactions and follows up with respondents for bank and administrative sources whenever a classification issue arises.

3.3.2 Other statistical procedures (e.g., data adjustments and transformations, and statistical analysis) employ sound statistical techniques

An example of sound estimation techniques was the 2001 survey constituting the basis for adjusting imports from a c.i.f. to an f.o.b. basis. In order to verify the coefficient used to adjust the cost of freight and insurance (initially 9 percent), the NBT conducted a survey with the support of the Goskomstat and the customs administration. The survey concluded that the correcting factor had to be increased to 13 percent, which was subsequently used to review current account data.

Estimates of shuttle trade combine information from transport and construction enterprises about construction material shipped in and out of the country. The NBT runs this survey once every two years and plans to widen the spectrum of surveyed entities.

3.4 Assessment and validation of intermediate data and statistical outputs

3.4.1 Intermediate results are validated against other information where applicable

The financial press is followed consistently, and information reported in it is used to identify high-value direct investment and other transactions for subsequent checking against reports. Moreover, press reports of major transactions involving the government are followed up to determine if full information about such transactions has been made available to balance of payments compilers.

3.4.2 Statistical discrepancies in intermediate data are assessed and investigated

The behavior of series is usually cross-checked with related series or indicators. Reported data by banks on investment income payments and receipts are regularly assessed in relation to investment flows as reported by the enterprise survey. Also, the comparison between trade in goods and trade data reported by banks allows deriving quantitative information on barter trade, on a merchandise by merchandise basis. The revised form to collect direct and portfolio investment was introduced in January 2004 and is expected to provide detailed breakdown for stocks together with valuation elements that will enable compilers to verify more thoroughly investment income, and investment flows against reported stocks.

3.4.3 Statistical discrepancies and other potential indicators of problems in statistical outputs are investigated

Net errors and omissions are analyzed to monitor developments in their size and sign when their magnitude reaches 10 percent of the total of credits and debits of the current account. Bilateral trade comparisons are frequently run with the four main-trade partners and have resulted in a joint exercise with Russia that led the NBT to revise its statistical treatment of workers’ remittances. At the present juncture, BOPSU staff do not use the Joint BIS-IMFOECD-World Bank statistics on external debt nor the BIS locational data for international banking statistics because of lack of training.

3.5 Revision studies

3.5.1 Studies and analyses of revisions are carried out routinely and used internally to inform statistical processes (see also 4.3.3)

The NBT conducts regular studies to take into account magnitude, classification, and orientation of revisions. These revision studies guide further consultations between the NBT and the Goskomstat, and they serve to compose the agenda and discussion programs with the Interagency Commission for Coordinating Balance of Payments, Preparation and Compilation, that are held twice yearly. Occasional discussions with the banking community and with administrative respondents help reassess data collection. The result of the revision studies serves to implement a medium-term revision cycle.

4. Serviceability

4.1 Periodicity and timeliness

4.1.1 Periodicity follows dissemination standards

Complete balance of payments statistics are disseminated quarterly, and their periodicity meets GDDS requirements. The NBT is bound by Decision 366 to issue these data in the prescribed periodicity and has done so without fail since it first started to publish quarterly data in January 1997.

4.1.2 Timeliness follows dissemination standards

The NBT releases data four months after the end of the reference quarter, with a timeliness consistent with GDDS requirements. Decision 366 imposes deadlines for publications that are strictly observed by the NBT.

4.2 Consistency

4.2.1 Statistics are consistent within the dataset

Concepts, definitions and classification are identical for quarterly and annual data. Financial transactions can be partly reconciled with those stock assets known to the NBT, such as the active balances of commercial banks with their nonresident correspondent banks, with a detailed breakdown by denomination currency. When its results become available, the revised investment survey is expected to provide by May 2004 similar detailed information for assets and liabilities of enterprises, regardless of their form of ownership.

4.2.2 Statistics are consistent or reconcilable over a reasonable period of time

Consistent time series are available since 1997; however, when methodological changes are introduced, only the former year is revised using the updated methodology due to constraints on capacity and the availability of source data. When changes in methodology are introduced, accompanying table notes are inserted in the graphics and tables, and a methodological note identifies and explains the main breaks and discontinuities. Unusual changes use references to exceptional operations that have been reported in the financial press and enable users to trace major financial news through statistical data.

4.2.3 Statistics are consistent or reconcilable with those obtained through other data sources and/or statistical frameworks

Balance of payments statistics are consistent with the other macroeconomic data: balance of payments are the source data used to assess resident and nonresident exchanges in national accounts by the Goskomstat. Consistency with government finance statistics is ensured since the MOF provides data on public publicly-guaranteed debt, as well as the nonguaranteed debt of public enterprises, together with payments of interest on the external debt. These data are incorporated in the balance of payments.

Regarding consistency with monetary statistics, this is observed, since the balance of payments source data for bank transactions with nonresidents are reported from information directly communicated by commercial banks. Monetary statistics are derived from the same financing and banking institutions, as compiled by the NBT Unit in charge of monetary and banking statistics from their balance sheet records.

4.3 Revision policy and practice

4.3.1 Revisions follow a regular and transparent schedule

The NBT does not release advance notice of revisions in its quarterly publications and does not implement a regular cycle of revisions. Quarterly publication may include revisions. Methodological documents are released when routine revisions are made to incorporate late or updated information, and when revisions are motivated by more substantial methodological and classification revisions.

4.3.2 Preliminary and/or revised data are clearly identified

Preliminary data are sent only to a limited number of users that participate in the decisions of the economic policy. Their preliminary character is clearly indicated on a cover note and in the tables. When preliminary data have to be disseminated again, there is a clear statement informing users that the preliminary data have been revised.

4.3.3 Studies and analyses of revisions are made public (see also 3.5.1)

Studies and analyses of revisions are limited to internal users and used as working documents reserved to balance of payments compilers. The result of studies may be shared with the Goskomstat and with the NBT authorities where there is a need to seek external support or a complementary view on the appropriate action that they entail. The NBT plans to share more of these revision analyses with a larger public.

5. Accessibility

5.1 Data accessibility

5.1.1 Statistics are presented in a way that facilitates proper interpretation and meaningful comparisons (layout and clarity of text, tables, and charts)

The quarterly data presenting the balance of payments are accompanied by tables, charts, and additional breakdowns, providing results in detail greater than the standard component of the balance of payments and for trade data. They are published in the quarterly publication The Balance of Payments of the Republic of Tajikistan and NBT Vestnik (News), in Tajik and Russian. Up until May 2003, BOP statistics were regarded as confidential and were not disseminated by the NBT; they served only the purpose of policymakers. Since that date, when BOP statistics are released to the public, this information is posted on the website of the NBT in a summary form, with a few notes in English, and a press release is disseminated to inform the press profession. In addition, more detailed data and more substantial comments are disseminated in an annual publication of the NBT, which is currently printed by BOPSU for the use of 26 registered users. This publication is available in Tajik, Russian, and English. The NBT plans to increase the visibility of this document and substantially enlarge its audience. For this purpose, in March 2004 the NBT has obtained from the Ministry of Culture its formal agreement that the balance of payments publication would become a registered publication of the NBT.

5.1.2 Dissemination media and format are adequate

Quarterly statistics are accessible on the Internet with an Internet browser, but it is currently not possible to download data on a spreadsheet. The NBT issues a press communiqué at the time when balance of payments data are officially released. Hard-copy formats are sent out to a handful of subscribers free of charge (e.g., Goskomstat library, the Dushanbe University) and can be consulted at the NBT documentation center located within the premises of the NBT. In addition, copies are disseminated to users that can explain that they use balance of payments statistics for professional reasons; for this purpose, the NBT requires a letter from the sponsoring agency.

5.1.3 Statistics are released on a preannounced schedule

The NBT does not maintain any official schedule for the release of statistics and relies on the regulation (which is a public document) that commands it to publish quarterly data within 90 days after the end of the reference quarter.

5.1.4 Statistics are made available to all users at the same time

The data are released simultaneously to all parties through the NBT website, in the monthly NBT Vestnik (News), and in the quarterly Balance of Payments of the Republic of Tajikistan.

5.1.5 Statistics not routinely disseminated are made available upon request

The NBT may provide customized tabulations to meet specific requests. However, the availability of additional statistics is known only to the working partners of the NBT, and this possibility is advertised during the meetings of the Interagency Commission for Coordinating Balance of Payments, Preparation and Compilation.

5.2 Metadata accessibility

5.2.1 Documentation on concepts, scope, classifications, basis of recording, data sources, and statistical techniques is available, and differences from internationally accepted standards, guidelines, or good practices are annotated

No metadata is made available to the public. At the occasion of the publication of the annual balance of payments statistics, the comments supporting tables provide some methodological information explaining the definitions and classifications used by balance of payments compilers. These comments are available in Russian and English.

5.2.2 Levels of detail are adapted to the needs of the intended audience

User needs for comprehensive metadata on sources and methods are not met.

5.3 Assistance to users

5.3.1 Contact points for each subject field are publicized

A general note is on the website and in the publications indicating that BOPSU is the contact source, with telephone and fax numbers, and email address. A specific contact person within the balance of payments area is not identified.

5.3.2 Catalogs of publications, documents, and other services, including information on any changes, are widely available

As yet there is no catalog of publications available; however, the NBT plans to release one by the end of June 2004 and, to that effect, has produced a mock version to test the format among NBT staff and a few external correspondents. In addition, the NBT maintains a display of current publications at its documentation center. The statistical documents, as well as their consultation, are free of charge.

Table 6.

Tajikistan: Data Quality Assessment Framework (July 2003): Summary of Results for Balance of Payments Statistics

(Compiling Agency: National Bank of Tajikistan)

article image
article image

APPENDIX I: Summary of the General Data Dissemination System

Data coverage, periodicity, and timeliness

Dissemination of reliable, comprehensive, and timely economic, financial, and socio-demographic data is essential to the transparency of macroeconomic performance and policy. The GDDS contains specific recommendations concerning coverage, periodicity, and timeliness for comprehensive frameworks as well as for data categories and indicators.

Quality

Data quality must have a high priority. Data users must be provided with information to assess quality and quality improvements. The GDDS recommends:

  • dissemination of documentation on methodology and sources used in preparing statistics; and

  • dissemination of component detail, reconciliations with related data, and statistical frameworks that support statistical cross-checks and provide assurance of reasonableness.

Integrity

To fulfill the purpose of providing the public with information, official statistics must have the confidence of their users. In turn, confidence in the statistics ultimately becomes a matter of confidence in the objectivity and professionalism of the agency producing the statistics. Transparency of practices and procedures is a key factor in creating this confidence. The GDDS, therefore, recommends:

  • dissemination of the terms and conditions under which official statistics are produced, including those relating to the confidentiality of individually identifiable information;

  • identification of internal government access to data before release;

  • identification of ministerial commentary on the occasion of statistical releases; and

  • provision of information about revision and advance notice of major changes in methodology.

Access to the public

Dissemination of official statistics is an essential feature of statistics as a public good. Ready and equal access by the public are principal requirements. The GDDS recommends; and

  • dissemination of advance release calendars; and

  • simultaneous release to all interested parties.

Plans for improvement

The GDDS recommends that plans for improvement be developed for all areas in which shortcomings exist and that these plans be disseminated.

The GDDS also recommends that any needs for assistance be identified in the metadata. This may also be helpful for donors and technical assistance providers to prioritize their activities.

For each participating member country, the GDDS metadata provide descriptions of the dimensions listed above, together with plans for improvement and needs for assistance. This information is posted on the Dissemination Standards Bulletin Board (DSBB); participating countries are encouraged to also post the metadata on their national websites.

Source: Guide to the GDDS, March 2002: http://dsbb.imf.org

APPENDIX II: Data Quality Assessment Framework—Generic Framework (July 2003 Framework)

article image
article image
article image

APPENDIX III: Users’ Survey

Summary of results of survey of data users

Local users of macroeconomic data have expressed interest in more details, higher frequency of dissemination, and enhanced timeliness for data on national accounts, government finance, and balance of payments. They noted the lack of information on dissemination of methodologies, advance release calendars, and data revisions information. A number of users have suggested that coordination among statistical agencies be strengthened. Some users noted that access to data was based on personal contact.

All users agreed that the ROSC Users’ Survey has formed a good basis for the future formal routine users’ survey by the Tajik authorities. The ROSC Users’ Survey shows that most respondents who use official data routinely are banks, universities, and research institutes, embassies, ministries and state agencies, and enterprises.

Table 7.

Questionnaire Results Analyzed by Type of User

article image
Source: [insert]
Table 8.

Results of Tajikistan’s User Survey

article image
article image
article image
NA = National Accounts; Prices refers to: CPI (Consumer Price Index) and PPI (Producer Price Index);GFS = Government Finance Statistics; Monetary = Monetary Statistics; and BOP = Balance of Payments Statistics