/ EUROPEAN COMMISSION
EUROSTAT
Directorate G: Business Statistics
Unit G-2: Structural Business Statistics /

Report on Eurostat Quality Visit to the Bulgarian Statistical Office

Date: / 2011-09-21
Version: / 1.0
Authors: / T. Courtel, S. Eeckels
Revised by:
Approved by: / Merja Hult

Commission européenne, 2920 Luxembourg, LUXEMBOURG - Tel. +352 43011
Office: BECH B3/411 - Tel. direct line +352 4301-+352 4301 34372

TABLE OF CONTENTS

Report on Eurostat Quality Visit to the Bulgarian Statistical Office - Page 1 / 9

Document Version 2.0 dated 08/02/2010

1. Purpose of the visit......

2. Institution......

3. Survey......

4. Relevance......

5. Definition of the target population......

6. Sampling Frame......

7. Sampling design......

7.1. Annual Survey......

7.2. Monthly survey......

8. Product nomenclature......

8.1. Overview......

8.2. Classification issues......

9. Method of data collection......

10. Confidentiality

11. Measurement errors......

12. Processing errors......

13. Non-response errors......

14. Data Processing and Quality Assurance......

14.1. Data Processing......

14.2. Quality Assurance......

14.3. Missing data......

15. Dissemination......

1.Purpose of the visit

The Eurostat PRODCOM team is conducting a number of "Quality Visits" to the PRODCOM teams in the NSIs, in order to discuss the methodology used, detect potential problems and identify best practice that can be shared with other countries.

After each visit, Eurostat will prepare a report on the discussions and, after checking it with the team visited, will load it to the PRODCOM CIRCA website, together with any material provided by the team in the NSI.

2.Institution

The Bulgarian Statistical Office is responsible for the collection and dissemination of official statistics. It consists of a Head Office and 28 Regional Offices. The PRODCOM team at Head Office (3 members) is part of the Business Statistics department, division Structural Business Statistics. Maintenance of the national product nomenclature is under responsibility of division “Statistical metadata, Classifications and Standards”.

The Regional Offices are responsible for collecting, entering and editing data.

3. Survey

The survey on production statistics in Bulgaria has been carried out since 1998, by means of which the industrial goods produced and the industrial services that have taken place on Bulgarian territory during the reference period (i.e. the calendar year) are surveyed.

The aim of the survey is to collect statistics for each product in the national Nomenclature of Industrial Production PRODPROM for which it is known that production and/or sales exists from enterprises manufacturing industrial goods and providing industrial services on the territory of Bulgaria. The survey covers enterprises which carry out industrial activity as primary or as secondary.

From 2001 to 2007 a national version of Prodcom List (PRODPROM) with national breakdowns and second measurement unitswas used.Since 2008 PRODPROM has been literally equal to the Prodcom List.

For each product manufactured the enterprise provides the following variables:

­Volume of Total production

­Volume of Total production on the basis of contract processing

­Physical volume of production sold

­Value of production sold

­Physical volume of production sold on the domestic market

­Value of production sold on the domestic market

­Total fee got by the subcontractor

­Fee got by the subcontractor from the foreign principal.

In addition, data on about 320 products is collected on a monthly basis.

4.Relevance

The main users of PRODCOM data can be presented in different groups, namely:

Classification of users / Description of users
United Nations
Institutions at European level / EUROSTAT
Multi-national organizations / UNECE, FAO,WBMS, International Fertilizer Industry Association etc.
Institutions at national level / Ministries (Ministry of economy, Ministry of Agriculture and Food, Ministry of Internal Affaires etc. )
Institutions at national level / Other governmental bodies, National Bank, other financial institutions, etc.
Trade associations, Embassies
Media (the main channel of statistics to the general public)
Researchers and students at international or national level / Scientific institutes, universities, students
Companies
Other statistics / National accounts, environmental statistics, SBS

5.Definition of the target population

All the enterprises that produce industrial goods and industrial services on the Bulgarian territory during the reference year are targeted by the survey.

6.Sampling Frame

The frame population includes all enterprises with their primary and secondary activities in sections B and C of NACE Rev.2 (excluding class 10.51. Data for products within this class is collected byMinistryof Agriculture and Food)which are surveyed by the exhaustive SBS survey. The SBS survey is the main source for drawing and maintaining the sampling frame of the Prodcom survey.

The period observed is one calendar year.

The observation unit is the enterprise.

7.Sampling design

7.1.Annual Survey

The PRODCOM survey is carried out as part of a yearly business statistical survey.

For the first time in 2010 ( for reference 2009 year) Prodcom questionnaire was included into Information System Business Statistics (ISBS), which provides online collection of annual reports of all economic active enterprises and contains a set of accounting and statistical questionnaires.

SBS survey covers all enterprises having carried out an activity in a given year. The use of an integrated Information System ensures that all the data is available on-line and can be correlated.

The sample design applied is a cut-off sample. The sample for 2010 survey includes all enterprises with total industrial turnover 120,000 Leva (about 62,000€) and over. The sample model ensures 90% coverage of sold production in terms of value at the national level for each NACE class.

To guarantee the required coverage, the threshold for a number of NACE classes is set at 20,000 Leva. The content of the sample is based on previous returns, and currently includes 9210companies with primary industrial activity, and 760 companies with secondary industrial activity.

7.2.Monthly survey

The monthly survey concerns 320 products at the 8-digit level.Information collected monthly for each Prodcom heading includes total production volume andvolume of production sold on the domestic market.

Data from the annual survey is cross-checked against the monthly data. The monthly survey meets the needs of international organisations, the Ministry of Agriculture and a number of other Ministries and Agencies. The products are selected based on

  • Their relevance for the National Economy
  • The interests of the traditional users.

8.Product nomenclature

8.1.Overview

Before 2008 the national version of Prodcom list, called PRODPROM contained national breakdowns and second measurement units.The codes were 10-digit. From 2008 the national product nomenclature is literally identical to the Prodcom List. The only exception is 47 headings for which a physical MU is used whereas Prodcom list requires only value.

PRODPROM is revised each year reflecting the latest changes and modifications made in the Prodcom List.

From January onwards, the PRODCOM list for the current year is used for the monthly data collection.

8.2.Classification issues

Enterprises classify their products, and can ask for assistance of the Regional Offices if required. The Accounting Service Bureaus that very often complete the Annual Survey on behalf of the enterprises have problems classifying the products. New products often result in classifications issues, and new companies lack the experience to classify their productscorrectly.

The NSI provides a web site with Frequently Answered Questions on classification issues.

9.Method of data collection

PRODCOM data collection is decentralized, organized through the Regional Statistical Offices. Data control is carried out separately at the regional level and at the national level.

The NSI of Bulgaria uses paper and on-line questionnaires.

The paper questionnaires:

•customized and dedicated to a certain activity at the four-digit level of NACE Rev.2;

•contain the entire list of products within the corresponding class.

•available on the NSI web site in Excel format; can be downloaded and printed.

Before 2009 the PROCOM survey was independent. From 2009 onwards Prodcom questionnaire is included in the Information System “Business Statistics” (ISBS). An agreement on a single entry point exists between the NSI and the National Revenue Office. The system is based on the experience ofPortugal.

The on-line system includes comprehensive consistency checking and validation of data, which makes it sometimes complex and slow to use, but the quality and completeness of the data is enhanced by the ability to compare Prodcom data with SBS data.

Companies are usually keen to complete the forms on-line because that allows them to provide tax returns on-line. After successful completion of the whole set of annual reports they obtain a unique ID code which is added to the Revenue Agency data base and allows to provide tax returns on-line.

The advantage of on-line data entry is instant data checking and feedback.

Respondents have an opportunity to make their choice:

•Submit all reports (incl. Prodcom questionnaire) in paper format to the Regional Office of the National Revenue Agency (along with tax returns).

•Enter the data on-line on the site of the NSI or the site of the Revenue Agency

About 50% of the data is submitted in paper format.

Paper returns are sent by the Regional Offices of Revenue Agency to the corresponding Regional Office of the NSI, where they are encoded using the same system used for electronic returns.

Large enterprises have to provide the information by the end of March, while small enterprises have until the end of April.

If respondents use on-line interface large enterprises can correct their submissions themselves before the end of March and small ones – by the end of April. After this deadline, only experts in the Regional Offices can correct the data.

Experts in the Regional Offices enter the data supplied on paper forms in the system until mid-July.

Data transmitted at the end of June are preliminary, and revisions are sent at the end of September and end of November.

About 22% of 2010 data was revised for the September transmission. A final transmission takes place at the end of November.

10.Confidentiality

The Bulgarian Statistical lawstipulates that the following data is confidential:

  1. individual statistical data;
  2. statistical data which can be matched in a way that enables the identification of a specific statistical unit;
  3. statistical information which aggregates data about less than three statistical units or about a population in which the relative share of the value of a surveyed parameter of a single unit exceeds 85 per cent of the total value of such parameter for all units in the population.

About 50% of the data collected is confidential, and about 40% of products are produced by only one or two enterprises.

11.Measurement errors

The classification of a product into a PRODCOM heading is made by the enterprise itself. In case the enterprise has doubts, it consults with the Regional Office. The expert suggests a heading code which has to be confirmed by the enterprise. The CN link is very helpful for enterprises in the classification process.

The main problems encountered during data collection are the following:

•Wrong industrial product code;

•Classification of new products;

•Wrong volume unit;

•Making a clear distinction between production and industrial services;

•Making a clear distinction between Sold Production and fee paid.

The on-line system includes numerous validity checks (including cross-checking SBS related data), but in many cases not entering PRODCOM data results in warnings only. This requires the Regional Offices to obtain the data (all methods to contact the enterprises are used).

The presence of SBS data is used to detect missing or erroneous Prodcom data (e.g. fee paid information) and vice versa.

12.Processing errors

A lot of micro- and macro-plausibility checks are performed at different aggregation levels (e.g. enterprise, NACE class, NACE division or PRODCOM headings) and by comparison with data of the previous year. In addition, outliers are detected using unit price.

Comparisons are made with SBS and STS data.

The integrated questionnaire uses a large number of cross-checking operations to ensure high data quality and availability.

An integrated database system is used to perform data processing and analysis.

13.Non-response errors

The response rate is quite good, due in part to the integrated system which allows entering taxation information.

Imputations are used to compensate forunit non-response. The information used for imputations includes historical data and data of similar units (by product and size class of the enterprise).

14.Data Processing and Quality Assurance

14.1 Data Processing

The Information System provides a number of arithmetic and logical controls, both mandatory and optionalbetween the items in each form and between the forms in the whole annual report. There are a number of arithmetic and logical controls inside the PRODCOM form.The System calculates both the totals and the unit price.

The Prodcom questionnaire can’t be completed successfully before all errors inside the questionnaire, detected by the checking program, to be corrected. The Prodcom data is cross-checked against SBS data and Prodcom data having been provided for the previous year.

If there are interlinked errors they are displayed after the validation of the whole set of reports.

14.2 Quality Assurance

Head Office uses the integrated database to produce a number of reports (e.g. per product – unit price) showing possible problems (such as outliers) with the data. Head Office then asks the appropriate Regional Office to verify the questionable data. As a rule, only the Regional Offices contact the enterprises. One of the problems related to data verification is that about 42% of the products are produced by one or two enterprises, which does not allow the determination of outliers. In these cases, the snapshots provided on CIRCA by Eurostat are very useful.

14.3 Missing data

The Regional Offices try to contact the enterprise assiduously by all possible means. If it is impossible to obtain the data from the enterprise, the data is estimated by using the appropriate means.

One of the disadvantages of on-line data entry is that the Regional Offices are not aware of possible problems with the data.

15.Dissemination

On the dissemination web site of the Bulgarian NSI, PRODCOM data is presented using the 6-digit CPA codes in order to decrease the number of confidential data. All variables (Sold Production in Volume and Value and Total Production in Volume) are published. Confidentiality rules are applied to each variable independently.

Additionally Prodcom data is contained intwo paper publications of NSI – Statistical Yearbook and Statistical Reference book. Another form of Prodcom data dissemination is against payment. The data is supplied per e-mail, on disk or paper.

1