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

Report on Eurostat Quality Visit
to Statistics Czech

Date: / 17.12.2010
Version: / 1.00
Authors: / Mihaela Agafitei
Revised by: / Brian Williams
Approved by: / Merja Hult

TABLE OF CONTENTS

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

2. Institution......

3. Survey......

4. Relevance......

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

6. Sampling Frame......

7. Sampling design......

8. Product nomenclature......

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

10. Confidentiality......

11. Measurement errors......

12. Processing errors......

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

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. The visit to the Czech Statistical Office is the 9th of these visits.

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 such as PowerPoint slides provided by the team in the NSI.

2.Institution

Official statistics in the CzechRepublic are organized and coordinated by the Czech Statistical Office ("CZSO").

3. Survey

The national Prodcom survey has been carried out since the 1970s, by means of which the industrial goods produced and the industrial services that have taken place on Czechoslovakian and subsequently on Czech territory during the reference period (i.e. the calendar year) are surveyed.

The Czech Prodcom survey is under the responsibility of the Statistical Department of Industry, Construction and Energy while the data collection and preliminary checks is under the responsibility of the Statistical processing department Plzeň.

4.Relevance

The main users are:

  • Government organisations
  • Public institutions and administrations
  • International organisations (Eurostat, UN)
  • The Business sector (business associations, enterprises, trade associations, etc.)
  • Others (e.g. researchers, students, media etc.)

5.Definition of the target population

All the industrial goods that are produced and the industrial services that are provided on Czech territory during the reference year.

6.Sampling Frame

In order to get information about industrial goods and services, the CZSO collects the required data from the enterprises. Therefore, the sampling frame contains all active enterprises which manufacture industrial goods and/or provide industrial services on Czech territory during the reference period.

The sampling frame is drawn from the Czech Business Register which is updated as follows:

Information source / Statistical Units / Frequency / Comments
Ministry of Justice / Business, Cooperatives, Foundations,
Physical Persons, etc / Daily
/ Only informational starting of new activity. So, it does not include information on the size of statistical unit.
Ministry of Justice / Business / Quarterly / Information on activity of the statistical units
Ministry of Industry and Trade / Business / Daily / Information on activity of the statistical unit
Social security system / Legal units having employees;
Self-employed persons / Monthly / Information on activity of the statistical unit
Statistical surveys via Coordination database / all / Depend on survey’s frequency / Varies information, manly related to activity; It ensures coherence between surveys.
Other registries / all / Vary / Varies information

In 2010 there were about 1.33 million active enterprises in the Czech Business Register of which there were 178,000 within the Czech manufacturing sector. The main errors found in the Czech Business Register are due to misclassification but it has an insignificant impact on the Prodcom survey since it occurs mostly for small enterprises.

The sampling and observational unit is the enterprise. The sample is drawn in December of the year before the reference period.

7.Sampling design

CZSO uses a cut-off sample (about 9,000 units) including all enterprises with 20 or more employees and having their main activity within the manufacturing sector. It is supplemented by non-industrial enterprises with an important share of industrial production. The selection of supplementary units is based on the experience of statisticians and overall knowledge of other surveys;

The survey frequency is annual, biannual and monthly. The monthly survey collects only total production and covers a very limited number of products and a limited number of enterprises in order to supply the necessary inputs to STS. The biannual survey satisfies specific national user’s needs. It is limited to Manufacturing and collects only total production. The annual survey collects all information and covers the whole sample. The biannual survey will be discontinued in 2011 due to budgetary reductions.

8.Product nomenclature

The national product nomenclature is based on the Prodcom list (8 digits) supplemented by some more detailed headings (9th and 10th digits) used for national purposes. Headings ending with 99 are used to indicate the production manufactured under subcontracting.

There are about 4510 headings of which 663 are outside Prodcom. The Prodcom volume unit is used in the Czech product nomenclature. The Czech List is always updated according to the Prodcom List (first eight digits always the same as the Prodcom List valid for the reference period) but the headings outside the scope of Prodcom are kept stable over several years.

CZSO has a strong interest in maintaining the time series and so it uses the last 2 digits to keep the List stable.

9.Method of data collection

Mailed paper questionnaires supplemented by electronic questionnaire (via internet, e-mail and CD-Rom). CZSO leaves to the enterprises the choice of method they use to complete the questionnaire. In 2010, 45% of responses were in electronic form.

The data is collected by Data Processing Centers located regionally, while a single Data processing center collects information on certain statistical domain from the whole country. In the case of Prodcom the data is collected by the Data Prosessing Center Plzeň, together with SBS and STS questionnaires. The workload is distributed by enterprise group. So a statistician is responsible for collecting all information related to a single enterprise, to make data collection easier and ensure a high coherence.

The survey timetable is shown below:

  • January Y-1 – General description of the Prodcom survey (The general methodology and questionnaire are approved. Possible changes are submitted for approval);
  • April Y-1 – Drafting the questionnaire
  • June Y-1 – Final questionnaire
  • September Y-1 – finalising the technical project
  • October Y-1 - Printing questionnaires
  • November Y-1 – final Czech product nomenclature is produced (PDF file, paper copies) together with correspondence and conversion table (e.g. Prodcom-CN; Prodcom T-1 – Prodcom T)
  • December Y-1 – drawing of the sample
  • January Y – Sending the questionnaires to enterprises
  • March Y – deadline for receipt of the completed questionnaires
  • March Y to June Y – reminders and data processing
  • June Y – first results of the surveys and data transmission to Eurostat
  • November Y – dissemination of the final Prodcom data

10.Confidentiality

The totals by product are confidential if

  • there are less then 3 contributors
  • one contributor contributes 85% or more of the total (in the past the share was 70% and the same percentage was used also for two dominant enterprises) Also, secondary confidentiality can result in a total being marked confidential.

The number of confidential totals by product is quite big and it creates problems in releasing a complete dataset.

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 provides the product’s description. The statistician 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

  • to classify the product when either no CN link exists, or the product can be classified according to CPA but no corresponding Prodcom heading exists, or in cases of contract processing or industrial services;
  • to report in the requested measurement unit;
  • to report production under subcontracting
  • some missing data.

The following actions are taken in order to minimise measurement errors:

  • During the whole data collection process, the staff maintains direct and close contact with the surveyed units. This is the case especially when a unit encounters problems in classifying a product to a Prodcom heading.
  • The enterprises are individually checked by the statisticians on the basis of their experience and information about the individual unit and the relevant economic branch.

12.Processing errors

A lot of micro- and macro-plausibility checks are performed at different aggregation levels (e.g. enterprise, NACE class, NACE division, CN or Prodcom headings) and by comparison with historical data. Also, outliers are detected using minimum/maximum/average price.

Comparisons are made with SBS and STS data,

The questionnaire is built so that some auxiliary information (e.g. the total value of production) is requested from the respondents in order to assess data quality

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

An Oracle system (in-house applications) is used to perform data processing.

13.Non-response errors

The average non-response rate is less 10%.

Report on Eurostat Quality Visit to Statistics Germany - Page 1 / 6