Production in industry

National Reference Metadata in Euro SDMX Metadata Structure (ESMS)

Compiling agency: Agency for statistics of Bosnia and Herzegovina 


Eurostat metadata
Reference metadata
1. Contact
2. Metadata update
3. Statistical presentation
4. Unit of measure
5. Reference Period
6. Institutional Mandate
7. Confidentiality
8. Release policy
9. Frequency of dissemination
10. Accessibility and clarity
11. Quality management
12. Relevance
13. Accuracy
14. Timeliness and punctuality
15. Coherence and comparability
16. Cost and Burden
17. Data revision
18. Statistical processing
19. Comment
Related Metadata
Annexes (including footnotes)
 



For any question on data and metadata, please contact: Eurostat user support

Download


1. Contact Top
1.1. Contact organisation

Agency for statistics of Bosnia and Herzegovina 

1.2. Contact organisation unit

Business Statistics Sector – Department for Industry and Construction Statistics

1.5. Contact mail address

Zelenih beretki 26, 71 000 - Sarajevo


2. Metadata update Top
2.1. Metadata last certified 14/06/2023
2.2. Metadata last posted 14/06/2023
2.3. Metadata last update 14/06/2023


3. Statistical presentation Top
3.1. Data description

Data on indices of the physical volume of production of industrial products are published for the section and division levels of economic activities in accordance with the Classification of Economic Activities of BiH 2010 (hereinafter: KD BiH 2010), for the Main Industrial Groups level (MIGs) and for the Total Industry level based on data collected through Monthly survey on industrial production (M KPS IND-1). 

3.2. Classification system

• Classification of Economic Activities of BiH 2010 (KD BiH 2010) for enterprises and local units (fully harmonized with all levels of NACE Rev.2).

• Classification of Products and Services of BiH fully harmonized with all 6-digit levels of CPA 2.1.

• More detailed Nomenclature of Industrial Products of BiH (Monthly NIP BiH 2018) – 8 – digit levels, identical to CPA 6-digit levels, and contains more detailed breakdowns to national aggregation of PRODCOM List 2010, updated with the PRODCOM List 2018

•  Main industrial groupings - MIGs 2009, in compliance with the definitions referred to in the Commission Regulation (EC) 656/2007)

3.3. Coverage - sector

Industrial coverage: An industrial production index covers the enterprises and their units employing at least 20 or more persons employed. Also, includes certain small industries with 10 or more persons employed within an industrial sector (sections B, C and D (except group 35.3) of the KD BiH 2010 (which is identical to NACE Rev.2)), mainly dealing with printing activities, bakeries, etc.). The units (parts of the enterprises) which perform industrial production (producing industrial products defined under the Monthly Nomenclature of Industrial Products of Bosnia and Herzegovina 2018 – Monthly NIP BiH 2018) and which operate within enterprises classified under the economic activities outside the referenced sections have also been covered.  This index has been calculated based on volumes and covers the reporting units which represent approximately 90% of value added at the 3-digit level (group level) of economic activities according to KD BiH 2010. It constitutes the coverage of approximately 1,600 reporting units for BiH.

Product coverage: This survey covers the products falling within the category of industrial activities and which are defined by the hierarchically structured classification of products, specified by the Nomenclature of Industrial Products – Monthly NIP BiH 2018, and updated with the PRODCOM List 2018, prepared for a monthly survey, which is basically grounded and identical to EU PRODCOM List. Production of products is covered regardless of whether they have been used by the reporting units for further processing/production or are intended to be sold. Both primary and secondary products (own and contracted production) have been covered. Produced quantities for the current month and cumulative quantities (from the beginning of the year) are collected on the product level.

Coverage by size classes: This survey covers all enterprises, natural and legal persons with at least 20 persons employed within sections B, C and D (except group 35.3) in accordance with KD BiH 2010 (NACE Rev.2). Units are selected in the sample by using cut-off sample. The threshold is lower for some groups of activities (with 10 or more persons employed) where the coverage would be lower than 90% in terms of the value-added proportion based on data obtained from the survey for Structural Business Statistics. 

3.4. Statistical concepts and definitions

The index shows the monthly movement of production in terms of the volume of production, that is, changes of the volume of production of products (goods and services) in accordance with the Nomenclature of Industrial Products for a monthly survey - Monthly NIP BiH 2018 which is harmonized with the EU PRODCOM List and which covers the products from the sections B, C and D (except group 35.3) according to KD BiH 2010 (NACE Rev.2). The Monthly NiP BiH 2018 comprises the products which have also been defined under the Classification of Products by Activity - KPPD BiH 2010 (CPA 2008), which implies the “product”, rather than “activity”, approach. Changes in the product quality relative to the changes in the quantity of production of products are taken into account when they actually occur. This is made by updating the weights. Concepts and definitions used in this survey are defined in the Methodological Document for calculation of a volume Index of Industrial Production which is available at: https://bhas.gov.ba/data/Publikacije/Metodologije/IND_01_2013_MD_0_BS.pdf  and in the methodological notes contained in the press release at the following https://bhas.gov.ba/data/Publikacije/Saopstenja/2023/IND_01_2022_12_1_BS.pdf

Detailed definitions of indicators are described in the new European Business Statistics (EBS) Regulation no. 2019/2152 and Commission Implementing Regulation (EU) 2020/1197 (Variable 140101 - Production (volume)) as well as in the EBS Manual 2021.

Through the Monthly Survey of Industrial Production (M KPS IND-1) data are collected from the reporting units in monthly periodicity, obtained at the level of industrial product, that is the quantity produced in the reporting month, and the cumulative quantity produced from the beginning of the year to the end of the reporting month, according to the measurement units defined in the Monthly NIP BiH 2018 (including the own production and/or contracted production). The Industrial Production Index (IPI) for BiH has been calculated based on the produced quantity of products which includes all products (goods and services) produced in the reporting month, whether sold or used as an input for further production or entered into the stocks of products for sale. Laspeyres 2-step formula is used. During step 1, indexes are calculated based on the quantity of every individual product (product level) defined by the Monthly NIP BiH 2018 and a relevant weight (that is, coefficient) which forms an estimated value added to every individual product. NOTE: Changes in the product quality relative to the changes in the quantity of production of products have been taken into account when they actually occur. After that, all indexes from the group level and higher aggregation levels are weighted with the shares of value added of every individual level of activities (group, division, section, MIGs, according to NACE Rev.2) of industrial production in total value added of the industry. Further explanations of the calculation method are provided in the Methodology which is available at the two foregoing links.

3.5. Statistical unit

Reporting units are enterprises and LKAUs dealing with the production of industrial products (goods and services) defined under the Nomenclature of Industrial Products. LKAUs are used in cases when the enterprise’s principal activity falls outside the NACE Rev.2 sections B, C, and D except group 35.3. This survey does not cover the reporting units (enterprises and LKAUs) outside the territory of Bosnia and Herzegovina). Full implementation of KAUs as reporting units, in accordance with requirements of new European Business Statistics (EBS) Regulation no. 2019/2152 and Commission Implementing Regulation (EU) 2020/1197, is done starting from January 2023. 

Observation units are all industrial products defined by Monthly NIP BIH 2018 produced by the reporting units.

3.6. Statistical population

The target population covers all active enterprises from sections B, C and D (except group 35.3) according to NACE Rev.2 and amounts to 6,591 active enterprises (note: entrepreneurs are not included in the target nor frame population) based on data available from the Statistical Business Register of BiH. Section B contains 252 units, Section C contains 5,969 units and 370 units are in Section D. Approximately 4,100 enterprises are those less than 20 persons employed in sections B, C and D (except 35.3) by NACE Rev.2.

Approximately 1,600 enterprises and KAUs are included in IPI calculation based on a chosen cut-off sample by using at least 20 persons employed threshold and in exceptional cases smaller enterprises (with less than 20 persons employed) in order to fulfill 90% of value added on group level of activity. The sample was drawn by using SBR in combination with data from PRODCOM and SBS surveys.

Statistical population includes the observation units – industrial products defined by the Monthly NIP BiH 2018 (harmonized with PRODCOM List 2018) which are produced by the reporting units (enterprises performing industrial activities, sections B, C and D – excluding group 35.3 or KAUs as a part of the enterprise which secondary activity falls outside the sections B, C or D except group 35.3 according to NACE Rev.2). 

3.7. Reference area

The BiH territory (includes two entities – the Federation of Bosnia and Herzegovina, Republika Srpska and the Brčko District)

3.8. Coverage - Time

Data on physical volume Indices of industrial production for BiH have been made (by using a single methodology) and published since January 2006 following a monthly frequency. They are available in the form of gross indices, calendar and seasonally adjusted indices, and trend-cycle indices

3.9. Base period

The base period is the average of the previous year. Weights are updated on an annual basis by using the chain-linking method. The reference year is also the previous year (currently, for the calculation of indices in 2022, the reference year is 2021).  Indices are published on the previous month, on the average of the previous year (chain indices), and recalculated to a common base year (2015=100) with chain linking. Series of gross, seasonally adjusted, calendar adjusted, and trend indices for the last 13 months are published in separate press releases (2015=100). The new base year 2021 will be implemented by the end of 2023. 


4. Unit of measure Top

Units of measure for every single product, which are used as inputs (product quantities) in the calculation of IPI are defined under the PRODCOM List – Monthly NIP BiH 2018.  This survey data has been published as the industrial production physical volume in the form of index numbers


5. Reference Period Top

The reference period is a calendar month.


6. Institutional Mandate Top
6.1. Institutional Mandate - legal acts and other agreements
  • Law on Statistics of BiH (Official Gazette of BiH No. 26/04 and 42/04), as well as laws on statistics of the entity's statistical institutions. The Agency for Statistics of Bosnia and Herzegovina is responsible for data collection, data processing and data publishing at the state level (BiH level) as well as for the implementation of methodologies in all three statistical institutions in BiH in a unique way;
  • Multi-annual statistical work programs for BiH (2021 – 2024) and Annual work plans of BiH;
  • Strategy for development of statistics of Bosnia and Herzegovina 2030;
  • Methodological guidelines for calculation of the industrial production physical volume index in BiH;
  • International recommendations and manuals (Eurostat, MMF, UN).
6.2. Institutional Mandate - data sharing

- Data delivery/sharing between the Institutes of Statistics from the entities (Republic of Srpska Institute for Statistics and the Institute of Statistics of the Federation of BiH) and the Agency for Statistics of BiH is regulated by Laws on statistics as well as Agreement on data sharing between all three statistical institutions;

  • Data sharing to external users -  state institutions in the country are regulated by Law on Statistics and other laws which regulate the work of state institutions (for example Directorate for Economic Planning of BiH, Ministry of Finance of BiH; Central Bank of BiH, etc.).
  • Statistical Office of the European Union (Eurostat);
  • External users – international organizations like UN Statistics Division (UNSD), International Monetary Fund (IMF), World Bank (WB), World Casting Metal Organization, etc. based on their requests and/or agreements signed;
  • Other users (students, academic community, individual users, etc. in accordance with special procedures defined by the Agency together with the individual agreements signed).


7. Confidentiality Top
7.1. Confidentiality - policy

Confidentiality of statistical data has been regulated by law, and personnel conducting statistical surveys are obliged, under the same legislation, to protect confidentiality.  The Law on Statistics of BiH (Official Gazette of BIH 26/04 and 42/04 - Chapter XI - Articles 23-29) sets forth the principle of confidentiality as one of the main principles.  The Agency of Statistics of BiH distributes statistics in accordance with the principles of the European Statistics Code of Practice, and especially with the principle of statistical confidentiality.  

7.2. Confidentiality - data treatment

The document titled “Book of Rules on Protection of Statistical Data in the Agency for Statistics” contains the procedure to secure confidentiality during data gathering, processing and dissemination - which includes protocols securing individual data to which access is allowed, rules defining confidential cells in the output tables and procedures for detection and prevention of subsequent disclosures, as well access to micro-data for the survey purposes.

Confidential data are not published and the minimum number of units for which data should be kept confidential in practice is less than three. In case of IPI this rule is not relevant because only indices are published. 


8. Release policy Top
8.1. Release calendar

The Release Calendar for Statistical Press Releases in BiH is published in advance for the entire year and it contains detailed information about all publications planned to be published. In December every year, the Agency for Statistics of BIH publishes on a website a calendar with accurate dates and times of publication of statistics for the next year.  The calendar is published on the website in four language versions (Bosnian, Croatian, Serbian and English languages) and in pdf format. IPI – on the 25th day of a month for the previous (reference) month (for example, 25th April for March as reference month). Release calendar for Statistical Press Releases in BiH for 2022 is available on the following link: http://bhas.gov.ba/Calendar/?lang=en

8.2. Release calendar access

Release calendar for Statistical Press Releases in BiH for 2022 is available on the following link: http://bhas.gov.ba/Calendar/?lang=en

8.3. Release policy - user access

All publications are published in accordance with the defined release calendar on exactly specified date, at 11 o’clock, on the Agency’s website.  If the publishing of the press release will be delayed, the relevant information shall be posted on the website and a new date/time shall be published.  All publications intended for users are available in 4 languages, that is, 3 local languages plus an English version. Users shall not have access to new data prior to their official publication on the Agency’s website, nor is it possible for them to be granted access to data earlier.

The principle that all users should have equal access to statistical data on an unbiased base is stipulated in Article 8, paragraph (2), subparagraph h) and Article 19, paragraph (2) of the Law on Statistics of BiH, and the Agency for statistics of BiH strictly comply with it in practice.


9. Frequency of dissemination Top

Industrial Production Index in BiH (IPI) is disseminated monthly.


10. Accessibility and clarity Top
10.1. Dissemination format - News release

Press releases of the IPI are published on the 25th day of a month for the previous month, in accordance with the defined date of publishing in the Release calendar, at 11 o’clock, on the Agency’s website: www.bhas.gov.ba in pdf format and also as a hard copy, upon request. The following press releases are hereby published:

 -  Industrial Production Volume Index in Bosnia and Herzegovina for the month of December 2022. 
https://bhas.gov.ba/data/Publikacije/Saopstenja/2023/IND_01_2022_12_1_BS.pdf

 - Monthly Industrial Production Indices for Bosnia and Herzegovina for 13 months, for example, December 2021 – December 2022”

https://bhas.gov.ba/data/Publikacije/Saopstenja/2023/IND_02_2022_12_1_BS.pdf

Indices are presented as gross, calendar, and seasonally adjusted as well as trend-cycle series for BiH for the Total Industry, sections B to D (except 35.3), MIGs, and divisions by NACE Rev.2

 
10.2. Dissemination format - Publications

 -  BiH in figures – Section “Industry” https://bhas.gov.ba/data/Publikacije/Bilteni/2023/NUM_00_2022_TB_1_EN.pdf

 
10.3. Dissemination format - online database

On-line database in Excel is available on the following link: https://bhas.gov.ba/data/Publikacije/VremenskeSerije/IND_01.xls

10.4. Dissemination format - microdata access

Micro-data are not disseminated and users have no access to them

10.5. Dissemination format - other

Data are transmitted to Eurostat which also publishes data for Bosnia and Herzegovina on its website

10.6. Documentation on methodology

Methodological explanations are available at the end of every IPI press release which is available on the Agency’s website:

https://bhas.gov.ba/data/Publikacije/Saopstenja/2023/IND_01_2022_12_1_BS.pdf

The Methodological Document for the calculation of an Index of the Physical Volume of Industrial Production in BiH is available on the Agency’s website:

https://bhas.gov.ba/data/Publikacije/Metodologije/IND_01_2013_MD_0_BS.pdf

 
10.7. Quality management - documentation

Development of the Quality Report is in progress and draft version is available internally on the intranet site of the Agency


11. Quality management Top
11.1. Quality assurance

Agency for statistics of BiH (BHAS) follows the recommendations for quality management provided in the European Statistics Code of Practice (CoP) and implements the guidelines provided in the European Statistics System Quality Assurance Framework (QAF).  More details about quality assurance and all procedures describing the BHAS’s quality policy are contained in the document:  BHAS Quality management policy and mechanisms available at website:

https://bhas.gov.ba/Content/Read/92?title=Kvalitet  

11.2. Quality management - assessment

IPI importance is primarily relevant to domestic and international users in terms of the country's economy quality assessment and it serves for GDP measurement in the National Accounts statistics (internally).  Data on IPI primarily result from the survey which receives a high response rate from the reporting units, as well as consistency of data provided through monthly surveys. IPI is calculated and published for BiH and transmitted to Eurostat in accordance with the methodology applied by all three statistics institutions in BiH in a harmonized manner.  Data are published on a regular monthly basis, accurately and on the same date, and are available to the users.  IPI is also EU and internationally comparable because it is developed in accordance with the EU and international concepts, definitions and classifications.  Also, it is comparable from the time distance and frequency of publishing, and it is comparable with other surveys, with SBS, PRODCOM and National Accounts in particular.

Note: In 2015, the Sector Peer Review for Business Statistics was conducted resulting in the survey quality assessment and the proposals for quality improvement.


12. Relevance Top
12.1. Relevance - User Needs

Key users of the survey in statistics data are:

- International main users: Eurostat, IMF, UN/UNECE, OECD, WB

- National main users: Directorate for Economic Planning of BiH with the Council of Ministers of BiH, Foreign Trade Chamber of BiH, Central Bank of BiH, Ministry of Finance of BiH, other bodies and institutions of the Government

-  Other users: business entities – enterprises, associations of employers, researchers and students, media and the general public.

Internally, these data are used for the National Account statistics for the production of the Gross Domestic Product and in the Structural and Short-term Business Statistics

12.2. Relevance - User Satisfaction

The Agency for Statistics of BiH is conducting a user satisfaction survey, carried out in 2020. This survey is conducted in 3-annual periodicity.  The total of 242 users were asked to respond to the questionnaire and 107 users provided their responses. Most interviewees visit the internet site and use the Agency’ statistical data occasionally (37.0%) or monthly (24.0%), at the same time, our website is daily visited by 7.0% interviewees. Most interviewees are interested in the statistical field of population statistics (64.0%), employment and salaries (56.0%). Report on the survey conducted is posted on the Agency’s website and is available to the users:

https://bhas.gov.ba/data/Dokumenti/Kvalitet/AZK2020_BS.pdf

Users’ main needs for data pertain to the website notification format. Namely, for the purpose of easier data management and their analysis, pdf format of the notification does not suit the users’ needs and efforts are being made for excel tables to also be posted.

Industrial production index in BiH is an important primary source of data to a large number of relevant authorities which are decision makers for the public and private sector. For example:
• enables the BiH authorities to prepare themselves for bilateral and multilateral negotiations within trade agreements,

• forms an important source of information for the Balance of Payments and National Accounts statistics and for economics studies,

• helps enterprises to conduct the market survey and develop their business strategy. 

12.3. Completeness

Having in mind the requirements of EU STS Regulation 1165/98 concerning the level of details for the IPI as well as length of time series, all these requirements have been met. The level of details for IPI time series are in line with the new European Business Statistics (EBS) Regulation no. 2019/2152 and Commission Implementing Regulation (EU) 2020/1197 (Variable 140101 - Production (volume)). Time series are available since January 2006 based on NACE Rev.2 for the Total Industry, sections and divisions levels as well as for the MIGs. 


13. Accuracy Top
13.1. Accuracy - overall

Considering a very high rate of responses, about 98%, overall accuracy is very good.  For all reporting units included in the sample, data must be completed before data compilation (reporting units are re-called or visited and, if even after re-calling it is not possible to obtain data from them, imputation shall be applied).  The exclusion of small enterprises shall not lead to an important bias in the series

13.2. Sampling error

There is no estimation available, nor has the sampling error been assessed, because it is very difficult to calculate it.  IPI results are calculated on volume data by applying cut-off sampling

13.3. Non-sampling error

For the purpose of this statistics, coverage has been analyzed annually and is made for the next year based on such analyses. It is not possible to quantify the coverage error because it concerns a sample that is not based on probability (random sampling). Considering a high response rate (approximately 98%), there is a small number of enterprises (small-sized mainly), approximately 2% of them, which do not provide their answers (unit non-response). In the case of unit non-response, the following imputation methods shall be applied: Based on data from the previous month or based on data on the production trend of the most similar units for the same product groups. In the case of non-response to certain items in the questionnaire (item non-response), reporting units shall be contacted by phone, until responses are obtained. Data are verified through logical and mathematical controls during the data entry process and after entering data into the IT application. A sample of a large number of small businesses (less than 20 persons employed) was not applied because of the reporting units’ excessive workload and the high costs of the survey.

Microdata are checked automatically during the data entry into IT application by applying numerical and logical checking controls as well as comparing data with data provided for the previous month). For each enterprise data are compared on the product level in order to find possible discrepancies. For the Entity territories, data are controlled by the Entity Institute/Office for Statistics, while for the territory of the Brčko District, the control is performed by the Agency for Statistics of BiH.

For the aggregated data, checking is done by different persons, firstly by persons directly involved in the calculation of indices and then by their head responsible for IPI. A lot of tables is prepared in Excel every month in order to check the data for each particular Entity and then to check the data/indices calculated for the BiH level. 


14. Timeliness and punctuality Top
14.1. Timeliness

Data collection has been done in a monthly frequency. Reporting units have the deadline for data delivery 15th in the month for the previous (reference) month. Eventually, two more days are available for re-calling of reporting units which didn’t deliver the data in order to get the data from all surveyed reporting units. The deadline for data publishing is T+25 days (for the level of BiH) for the Preliminary results. Preliminary results shall become final on the date of publication of the press release for the next month.

14.2. Punctuality

In general, there is no difference between the announced and actual date of the publication (according to the Release Calendar). All Press releases are published on the date of publication as foreseen in the Release Calendar.   


15. Coherence and comparability Top
15.1. Comparability - geographical

Considering that the IPI results for BiH aare published in accordance with the EBS Regulation 2019/2152, the results are comparable with those of other countries (EU primarily) which apply these regulations, methodology, and relevant classifications.

15.2. Comparability - over time

The industrial production index for BiH has been calculated and published since January 2006 in a monthly periodicity by using the NACE Rev.2 classification at the Total industry level, section level, division level as well as at MIGs level. In 2013 all IPI series that were calculated by using NACE Rev.1.1 are backcasted on NACE Rev.2 by using the macro approach (on a macro level) as well as rebased from 2005=100 to 2010=100, and since the beginning of 2018, the base year was updated from 2010 = 100 to 2015 = 100, in order to ensure comparability.

15.3. Coherence - cross domain

Consistency with the annual industrial production indicators is achieved with:

- the annual PRODCOM survey in which data have been collected from a larger number of reporting units (cut-off sample of reporting units – enterprises and KAUs with at least 10 persons employed), by using the NIP BIH/PRODCOM classification which is annually updated with the official PRODCOM List.  It is possible to compare a product to a product. Currently, NIP BiH/PRODCOM 2021 is using PRODCOM List 2021. PRODCOM results are used for the calculation of weighting coefficients on the product level for the base year (currently 2015). 

- SBS survey results are used for the calculation of the weighting structure (value added) by factor costs which are available at the group level (3-digit level) and the higher levels of aggregation according to NACE Rev.2. SBS data are used for the calculation of weighs for weighting the industries for the groups, divisions, sections and MIGs levels by NACE Rev.2. Consistency of IPI data with the SBS and PRODCOM survey data is verified every year.  

15.4. Coherence - internal

Approximately, it is also possible to compare this survey data with the National accounts and, in most cases, data are consistent. The Laspeyres index which is annually linked through the chain-linking is used for the calculation of the IPI BiH.  Higher levels of IPIs are annually linked by using annual overlapping, that is, an annual average of the year t-1.  IPI is calculated by applying the chain-linking method for the levels of the classes, groups, divisions, and sections of economic activities, and also for the level of the Main Industrial Groupings (MIGs) as well as the Total Industry level of the NACE Rev.2 classification (KD BiH 2010).  IPI for BiH is also adjusted by using TRAMO/SEATS method based on ARIMA models. Adjustments are done for the same Total Industry level, section, and division levels as well as for MIGs. Adjusted data are checked and compared with gross data and they are coherent. Coherence is achieved by compilation methods and a weighting system. IPI for various levels of the NACE is calculated in two stages according to the Laspeyres formula


16. Cost and Burden Top

Currently, data on industrial production used for IPI calculation are collected by using paper questionnaires which requires a lot of money and human resources. Data are collected by three statistical institutions: Republic of Srpska Institute of Statistics for the territory of Republic of Srpska, Federal Institute of Statistics of FBIH for the territory of Federation of BIH and the Agency for Statistics of BiH for the Brcko District of BiH territory. Approximately, 23 people are involved in data collection process and data compilation process on annual basis in full-time equivalent. In addition, one IT expert per institution is responsible for development and maintenance of IT application (30% of their working time). All three statistical institutions have very high costs of printing the questionnaires, postal and telephone costs.

Response burden is not very high, approximately 25 minutes for fulfilling the questionnaires per respondent.


17. Data revision Top
17.1. Data revision - policy

There is no officially defined revision policy except in the case of errors in the data, wherein the corrected data shall be marked as "corrected data" and accompanied by an explanation to all users of the reasons for the data correction. This is published in the official publication as well as on the website of the Agency. After publishing corrected data on the website of the Agency, these data are also transmitted to Eurostat.

17.2. Data revision - practice

Usually, big revisions are done in a way that representatives of all three statistical institutions have to discuss together and agree on new proposed Statistical Standards/methodologies which will require data revision. Main changes in the methodology, a change of the base year and similar shall be published on the Agency’s website.  All methodology changes are thoroughly explained in the Press release, in the section Methodological Notes.

Data revision, which occurred by publishing seasonally adjusted data for the previous months, is not published to users, except in the Methodological notes in the IPI Press release. These notes contain a description of the applied method of seasonal adjustment (the scope of the data revision incurred by applying seasonal adjustment are not measured).  Corrections of Press releases resulting from errors shall be published automatically by releasing a revised/corrected Press release including a note and explanation to users concerning the reason of the revision/correction. Revised/corrected data are also transmitted to Eurostat on the same date when revised/corrected data are published on the website of the Agency.  IPI data are published every month for the previous (reference) month as preliminary data, including a note that, when publishing IPI data for the next month, preliminary IPI data from the previous month shall become final.  IPI is published on the 25th day of a month unless the 25th day falls on a non-working day or weekend day.  In such a case, IPI data shall be published on the first working day to follow, as defined in the Release Calendar. Regarding the level of revision, gross (unadjusted) and calendar-adjusted indices are revised rarely with minor changes in the level of indices in last 36 months. Seasonally adjusted indices are regularly monthly revised due to seasonal effects with no significant influence on the level of revision. The Agency for Statistics of Bosnia and Herzegovina doesn't calculate quantitative indicators of revisions yet.   

Mean Revision (MR) and Mean Absolute Revision (MAR) and are used as revision indicators. They are calculated on year-on-year growth rates for original data. The calculated values for the last 36 months monthly IPI index are: MR= -0.054 and MAR=0.674. 


18. Statistical processing Top
18.1. Source data

Monthly statistical survey: Monthly survey on Industrial Production (M KPS IND-1) is the main data source for IPI calculation, which is conducted in a traditional way by sending out paper questionnaires to the field by mail. The survey covers approximately 90% of the total Value Added of the industry on 3-digit levels of NACE Rev.2 (KD BiH 2010). 

18.2. Frequency of data collection

Data collection frequency is monthly, with deadlines 15th day in the current month for the previous (reference) month during 2022.

18.3. Data collection

IPI data for BiH are collected through the M KPS IND-1 statistical survey conducted by mail and using paper questionnaires.  Data are collected through the cantonal/ regional offices of the Entities – for the Entity level and through the Brčko District Branch Office of the Agency for statistics of BiH for the territory of the Brčko District

18.4. Data validation

The Entity Institutes/Offices for Statistics perform entry and validation of data for their Entity territory, while the Agency performs that for the territory of the Brčko District.  The Entity Institutes/Offices provide the Agency with all data required for the IPI compilation at the BiH level, in agreed formats.

Data cross-checking controls are applied. If “Suspicious” data are found, they are sent back to the Entity's statistical institutions for checking and verification.  The Entity Institutes/Offices for Statistics recheck data and, if required, a phone call-back procedure is applied or pays a visit to them, if it is necessary.  This is a standard procedure that has been applied for a long period of time, enabling the Industry Statistics departments to obtain data of good quality.  Data validation is done automatically by applying predefined data controls through IT applications. The validation process includes the following: cross-cutting, comparison with previous data sets, etc.  Afterwards, indices are produced and results are also cross-checked with historical data.  All verifications are performed through the IT application for Industry.  For the data transmission to Eurostat, a file format has been developed in IT application. A new SDMX-ML format is used for data transmission to Eurostat. No validation controls and checking are applied to this file because the national validation rules have already been applied.

18.5. Data compilation

IPI for BiH is Laspeyres type chain index which is calculated based on coefficients/weights from 2015 (as a base year) at the product level, while the weighting structure from the previous year is used for the activity structure (activity level). 

Index compilation has been done as follows:  it starts from the micro level – product level where the total production is calculated for every product defined under the Monthly NIP BiH 2018 through the aggregation of all productions of the same product across all producers.  Thereafter, these quantities are multiplied by the weighting coefficient in order to get the value-added “mass” for each product, which is then summed to the 4-digit level of KD BiH 2010.  Then, the “mass” are summed from 4 to 3-digits (Group levels) of KD BiH 2010. After that, relatives are calculated from this “mass”. Afterward, these relatives are multiplied to the 2-digit levels of KD BiH 2010 and higher levels as well as for the MIG level, by using coefficients/weights of value added obtained from the SBS survey.  Basically, the IPI calculation consists of 2 steps:

Step 1 – indices at the 4- and 3- digit levels of KD BiH 2010 are calculated from volume data obtained by multiplying the produced quantity of each product defined under the Monthly NIP BiH 2018 and appropriate weighting coefficients, using the Laspeyres formula.


Step 2 – for all indices of higher level and MIGs, they are calculated by aggregating the group indices (3-digits) of KD BiH 2010 - weighted with the “Group” share of Value Added in the total Value Added.

 

Coefficients/weights used for every individual product defined under the Monthly NIP BiH 2018 are updated on a 5-year basis.  Weighting coefficients are calculated based on the product’s unit value resulting from the PRODCOM BiH 2015 results and estimates of value added which is obtained from the SBS survey for 2015 on the 3-digit level of KD BiH 2010.  

Weights for activities from the 3-digit level of KD BiH 2010 upwards are calculated from the gross value added at factor costs from the SBS survey which is corrected by the Industrial Production Physical Volume Indices, in order to obtain an approximation of the previous year’s value added.  These weights for the activity level (activity structure) are updated every year

18.6. Adjustment

Seasonal adjustment of gross IPIs has been done by using an automated model through software JDemetra+ v2.2.3 and TRAMO/SEATS methods (regression method), which is based on the ARIMA models, and indices are published as gross, seasonally and calendar adjusted and trend indices. The direct method is used in the Entity statistical institutes while for the BiH level, adjusted indices from the entities are used and multiplied with fixed weights over five years (share of each Entity value added in the total BiH value added).  



Annexes:
Seasonal Adjustment_BA


19. Comment Top

No comments


Related metadata Top


Annexes Top
SA_BA