Turkish Statistical Institute (TurkStat)

Short-term Business Statistics Department (Industrial Statistics Group)

Turkish Statistical Institute (TurkStat) Devlet Mah. Necatibey Cad. No:114

06100 YUCETEPE/ANKARA/TURKEY

Industrial Production Index

NACE Rev. 2.

Mining and Quarrying Sector (B), Manufacturing industry (C) and Electricity, Gas, Steam and Air Conditioning Supply (D) classified in NACE Rev.2 are covered.

35.12, 35.13, 35.14, 35.2, 35.3 and 09 sectors of NACE Rev.2 are not covered.

Reporting unit : Local unit and enterprise

Observation unit : The Industrial Production Index covers the mining, manufacturing and electricity, gas, steam and air conditioning supply industries (categories B, C and D of Nace Rev. 2).

Enterprises, whose main activity is mining and quarrying (B) or manufacturing industry (C) or electricity, gas, steam and air conditioning supply (D) are covered. 

The Industrial Production Index with 2015=100 reference years is calculated with the production values from 6.452 (by April 2023) enterprises and the production value estimates obtained from the turnover of more than 300 thousand firms from the Revenue Administration.

Whole of Turkey.

Monthly survey was being implemented since 2005 .

The publication calendar is set about 40. day (t + 40) after the end of the reference month.

Reference year : 2015

Indices and percentage changes

Month

The legal basis is the Statistical Law no. 5429, of 10th November 2005. The obligation on units to provide data is done via compulsory surveys. Fines will be given in cases of non-response or deliberate reporting of wrong information according to the Statistical Law no. 5429. There is no planned changes in legal basis, obligation to respond and frame used.

Unadjusted, calendar adjusted and seasonal and calendar adjusted indices are transmitted to Eurostat. If requested indices are shared with all international and national instutions and researchers under data confidentiality rules.

All statistics collected and published by TurkStat are governed by the Turkish Statistical Law no. 5429, of 10 November 2005.
Under the terms of the Statistical Law:
1. TurkStat has the independence to determine the methodology and release dates of statistics for which it is responsible.
2. Data are collected only for statistical purposes and confidentiality is assured.
3. The confidentiality of individual responses is guaranteed.

Confidential data can be accessed only by those involved in the production of official statistics, to the extent that they need for performing their duties properly. If the number of the statistical unit in any cell of the data table formed by aggregating the individual data is less than three or one or two of the statistical units are dominant even if the number of units is three or more, the data in the concerned cell is considered confidential. The confidential data are compiled, processed and preserved for the production of official statistics cannot be delivered to any administrative, judiciary or military authority or person, can not be used for purposes other than statistics or as an instrument of proof. Civil servants and other staff in charge of compiling and processing these data are obliged to comply with this rule. This obligation continues after the related personnel leave their duties and positions. The rulers of the institutions and organizations producing official statistics shall take all measures to prevent any illicit access, use or revelation of the confidential data. Data or information obtained from sources that are open to all people shall not be deemed confidential.

Data confidentiality ceases when a statistical unit gives written approval for the revelation of confidential data concerning itself.

Principles and procedures relating to data confidentiality and security shall be regulated through statutes to be issued in line with national and international principles and by soliciting the opinion of relevant institutions and organizations.

Note: The above information is regulated in Articles 13, 14 and 15 of law no. 5429.

At the end of each year TurkStat publishes a release calendar providing the release dates for the coming year. This is also disseminated on TurkStat’s internet website (www.turkstat.gov.tr).

Any major changes in methodology are announced at the time of the change and the data compiled using the revised methodology is released.

Summary information regarding the revision policies of statistics to be published within TurkStat in 2023 is given below link by topics.

https://www.tuik.gov.tr/Kurumsal/Cari_Yil_Revizyon_Plani

Release calender link  is given below.

https://www.tuik.gov.tr/Kurumsal/Veri_Takvimi

The data are released simultaneously to all interested parties by issuing the press release Industrial Production Index.

The press release is prepared and kept secure under locked conditions until the specified time of release, at which time it is made available to the media and all interested parties.

Simultaneously, the press release is posted on the TurkStat website www.turkstat.gov.tr 

Identification of internal government access to data before release : no users have prior access to the data before its general release.

Transmission to Eurostat and further use of the statistics : At the time of national publication unadjusted data as well as working day, seasonally and calendar adjusted data are transmitted to Eurostat via eDAMIS using SDMX (xml). The data have been sent in SDMX format to Eurostat since 2016.

Monthly  (Within 40 days after the end of the reference month)

The data are released simultaneously to all interested parties by issuing the press release Industrial Production Index. The press release is prepared and kept secure under locked conditions until the specified time of release, at which time it is made available to the media and all interested parties.

Simultaneously, the press release is posted on the Turkstat website www.turkstat.gov.tr

• Data are disseminated on the Industrial Production Index by using Chained Laspeyres index (2015=100) since March 2018. The index covers the mining, manufacturing, and electricity, gas, steam and air conditioning supply. 
• The data are presented by main economic activities according to the NACE Rev. 2 classification, at the two-digit level. 
• The index is re-based and the weights are revised. 
• The data are seasonally and calendar adjusted at total index, section level, Main Industrial Groupings (MIGs), technology classification and two digit economic activities level. Seasonal adjustment and calendar adjustment for Industrial Production Index is made by indirect method. 

Online database is available. Online database can be accessed via https://biruni.tuik.gov.tr/medas/?kn=67&locale=en

The data are released simultaneously to all interested parties on the Internet website of TurkStat issuing a Statistical release at the announced release calendar. A notification about the release is simultaneously mailed to the media and other subscribers of the news service via social media.

Micro-data access is not available to the public or private users.

Unadjusted, calendar adjusted and seasonal and calendar adjusted indices are transmitted to Eurostat.

Methodology and sources used in statistical compilation are disseminated on the Turkish Statistical Institute (TurkStat) website http://www.turkstat.gov.tr

The monthly press release provides a technical explanatory note about the data. In the web site near press release short metadata can be accessed. Also all statistical tables include metadata.

Criterias for quality assessment in Turkish Statistical Institute are found at 

https://www.tuik.gov.tr/indir/m_t_metaveri/sue_mv.pdf  (in Turkish only)

Quality checks and validation of data are done throughout the whole compilation process.

Internal consistency checks are carried out at the analysis stage. As a part of the review process, changes in the Industrial Production Index are compared with Turnover Index and National Account Statistics.

Industrial Production Index are detected for below criteria;

- Relevance; enterprises are checked, which indices are calculated from, main activities with SBS, BR and other surveys. Opinions from users (public or private) are sought on the calculation of indices, and technical support/advice is also sought from Eurostat staff.

- Accuracy and reliability; Data source and results are regularly assessed and validated. Revisions to develop statistical process is analyzed regularly.

- Timeliness and punctuality; Dissemination of press release and indices day and time announced to public at the beginning of the year. Can be accessed here. https://www.tuik.gov.tr/Kurumsal/Veri_Takvimi  Possible deviations from the publishing calendar, and a new date is notified to the public.

- Comparability; Production index is comparable with EU and other countries. Methodological studies are done with close cooperation between Eurostat.

- Coherence; Eurostat and other international standards are covered.

Indices are based on reliable data; Publication of results are always punctual; Multiple plausibility checks with different sources are done.

Short-term statistics are used by different national and international institutions and non-institutional users:

Institutions: Ministries, National Central Bank, political institutions (e.g. parliament), local authorities

International institutions: Eurostat, OECD, UN, IMF

Social actors: Trade unions

Researchers and Educational Institutions

Media

Enterprises/businesses

Internal users

Public

The user satisfaction survey is implemented by TurkStat. However this survey is not applied especially to industrial statistics. There are regular meetings with national experts and essential users to react to new requirements. 

The production index is compliant with STS-Regulation.

Questionnaires (Industrial Production Questionnaire) for the reference month are filled by enterprises between 11^th-30^th of following month. Forms filled out by the  enterprises by using internet base programme. The survey response rate is approximately 99.9%. Non-responding units are followed up on telephone, e-mail or by interviewers. When a unit refuses to respond, the unit is reminded of its legal obligation to respond. In case of no response, the enterprise' turnover change rates are used (using VAT data).

Index numbers are revised due to late response regularly and occasionally due to the change in statistical methods, concepts, definitions, classifications and the base year.

In the unadjusted series;

1) When publishing the current month indices previous month indices will be revised

2) In the February press releases, previous year indices will be revised.

For the Seasonal and Calendar Adjusted Series;

1) Revisions  are implemented on the data of the last three years excluding the current year. As the IPI is published with base year 2015, as a special case only applicable to 2018, revisions are implemented on the data of the last two years excluding the current year when seasonally and calendar adjusted series are published.

2) At the end of the year the model structure of the series, model parameters and the outliers observation structure will be changed if the method requires change.

First estimations are consistent with the revised indices calculated at the end of the year. 

Data source and results are regularly assessed and validated. Revisions to develop statistical process is analyzed regularly. 

Currently no calculations on sampling error are available for the Industrial Production Index on any level.

Response rate

Industrial production data were collected only by means of questionnaires till 2018. The 2010 = 100 based Industrial Production Index, which is published between 2013 and 2017, was calculated using survey data from approximately 10 thousand companies. As of March 2018, Revenue Administration microdata has been used to produce more comprehensive data sets, to reduce respondent burden and to make public resources more efficient. The Industrial Production Index with 2015=100 reference years is calculated with the production values from 6.452 (by April 2023) enterprises and the production value estimates obtained from the sales of more than 300 thousand firms from the Revenue Administration in 2023.

Sampling error is not calculated for the survey.

There is no coverage problem in the frame. 

The survey is compulsory. The response rate for industry is approximately 99%. A lot of checks are carried out at the micro and macro data level. For example the following processing errors may occur:

Human errors during the data editing process: these errors are often recognized with automated checks imputation errors: the imputed value is not within a plausible range plausibility errors: a lot of checks are carried out automatically e.g. check of completeness /  sums / average rates / minimum-maximum values,…. The qualified checking of original respondent data, contacts to reporting units for corrections and the usage of sophisticated plausibility routines shall minimize the processing errors on micro-data-level.

Final checks with analyze-tables (aggregates/more complex relations). The macro data tabled are verified by computer programs but also by the staff whether these macro-data as publication figures look plausible. The persons responsible for the index calculation have their own computer programs to test the index-plausibility of the different data at 4- and 2-digit-level and for all aggregates. If inconsistencies are detected the data must be verified at micro-level again by using the raw data of the questionnaire.

Actions to speed up or increase the rate of response

Non-responding units are followed up by telephone contacts, by e-mail or visits by field officers.

The indices are calculated monthly and disseminated in t+40 (average) days. All aggregation levels and sub-sectors are released at the same time. The results of the  production index are released on the 9-10th day of the second month following the reference month. The data are released simultaneously to all interested parties by issuing the press release “Industrial Production Index”.

The questionnaires of each reference month are filled by enterprises about 11 days after the reference month.

The news releases are published according to the announced release calendar.

Industrial Production Index is comparable with EU. Methodological studies are done with close cooperation between Eurostat. The same statistical concepts are applied in the entire Turkish National territory. No geographical discrepancies exist.

Data on industrial production are comparable over time. When significant changes are introduced (e.g. change of classification, base year or methodology), historical time series are recalculated as far back as it is reasonably possible. Time series according to NACE Rev. 2 are available from 1986 onwards.

2015=100 Industrial Production Index is calculated by the chained Laspeyres method, which is chained using annual averages. With the transition from fixed base year index to chain index, the most recent weights are used every year. The value-added weights (t-1) obtained from National Accounts and Structural Business Statistics are used. With the new data source and the structure, 2015=100 series have been calculated as from 2010. Between 2005 and 2010, the annual average change rate are used to link the new series.

The Industrial Production Index can be compared with the following data;

Quarterly national account data, quarterly labour input indicators, tendency survey, turnover index, annual national accounts data, annual structural business statistics, annual industrial product statistics and annual retail trade statistics. The coherence between IPI and national accounts is high.

Cross-checks with;

• absolute STS data in industry 
• compare over period: compare with previous STS-values (previous month and year)
• SBS data (e.g. activity, turnover)
• Other surveys e.g., foreign trade statistics and tendency Survey

Statistics on cost and burden are available for the Industrial Production Index. A one-time survey was implemented businesses for measurement of workload in 2015 and 2019. Statistics on cost and burden are not available for the year 2023.

Information about routine/major revisions to be implemented in the current period is crucial for statistics users in the sense of determining the validity of data. Because of changes in the reply of respondent unit and procedures such as seasonal and calendar effects adjustment, data might be revised routinely. This information is important for the validity of data in time. Summary information regarding the revision policies of statistics to be published within TurkStat in 2022 is given below link by topics.

https://www.tuik.gov.tr/indir/revizyon/2022_RIF_IPI_EN.pdf

Instruction on methods and principles regarding revisions on statistical data produced by Turkish Statistical Institute is given below link;

https://www.tuik.gov.tr/Kurumsal/Yonergeler (Directive on Methods and Principles Regarding Revisions on Statistical Data Produced by Turkish Statistical Institute)

Data source: Data used for calculation of the index are consist of Monthly Industry Production Questionnaire data and turnover data calculated from Revenue Administration data base. The published data is temporary and data for the previous month is revised every month. 

In addition data are obtained from Turkish Electricity Transmission Company (for electricity) and Tobacco and Alcohol Market Regulatory Authority.

Nature of weights: Up to the level of the section according to NACE Rev. 2 classification (2 digit level) weights obtained from the Structural Business Statistics are used. For the upper level indices National Accounts value added is used as a weight. For each level t-1 periods weights are used.

Variables: Production value, turnover, production quantity and number of days worked 

In the index calculation the production quantity is used in sectors;  05- Mining of coal and lignite, 06- Extraction of crude petroleum and natural gas, 12- Manufacture of tobacco products and 35.11- Production of electricity. Total number of days worked obtained from the data of Social Security Institution is used in sectors; 30.11-Building of ships and floating structures, 3012- Building of pleasure and sporting boats and 30.20- Manufacture of railway locomotives and rolling stock. In the index calculation deflated production value and turnover is used for other sectors.

Deflator: Domestic and non-domestic producer prices

Type of index: Chained Laspeyres Index method (weighted) is used.

Classification: Nace Rev.2

All data is collected on a monthly basis.

Data used for calculation of the index are consist of Industry Production Questionnaire data and turnover data calculated from Revenue Administration data base. 

For the survey data are collected by web-based data collection method. Non-respondents are contacted primarily through phone and e-mail.

During the production phase enterprises are contacted in order to get data from important and/or larger enterprises. Contact is primarily through phone and e-mail. The survey has a sanction system with a penalty for not responding.

The data entry control of the questionnaires is performed by the Regional Offices via a web based computer software. The software for the data entry is designed to check the consistency at several points and not to allow entering wrong or missing data. 

Appropriate validation procedures are applied during the whole production process of statistics, and they are in line with rules provided by Eurostat, i.e. starting from selection of respondents to compiling indices and transmission to Eurostat. Data validation system incorporates many tools for data control, estimation and imputation, e.g. programme procedures applied for error checking and detailed reports on data completeness helpful in analysis of non-response. Validation procedures are applied on the level of individual respondents and on aggregated level as well as before data transmission to Eurostat. Data validation on the level of individual respondents is based on: special programme procedures used for errors checking when data are inserted to the system these procedures help to avoid errors at the stage of data collection what is essential for final data results. information given by an enterprise on the reasons of substantial difference in the level of data reported in the current period as compared to the equivalent data for the previous period. 

In the case of any doubts about data correctness there is a direct contact with an enterprise in order to explain outliers. Both the values and dynamics are compared with previous month and corresponding month of the previous year. Validation procedures are also performed before data transmission to Eurostat. The structured data files in SDMX format are created and data assigned to the codes are checked in detail. In case of significant data revisions, for example when the base year changes or seasonal adjusted data are revised, the whole time series are submitted to validation procedure.

Estimates for non-response

There is rigorous follow-up of nonresponding units. If enterprises refuse to respond, the unit is reminded of its legal obligation to respond. But if the problem persists, imputation process is applied for the unit. In non response situation, the annual rate of change of the sectors are used for imputation.

Estimates for grossing-up to population levels

Sample results are not grossed-up.

Type of index

Chained Laspeyres Index method (weighted) is used.

Method of weighting and chaining

Up to the level of the section according to NACE Rev. 2 classification (2 digit level) weights obtained from the Annual Industrial and Service Statistics are used. For the upper level indices National Accounts value added is used as a weight. For each level t-1 periods weights are used.

Planned changes in production methods

TurkStat do not plan any changes for the production method.

Seasonal adjustment: 

Seasonal and calendar related effects prevent observing the general tendency of data because of their temporary characteristics. Therefore, identification of seasonal patterns of short term indicators plays crucial role in order to make reliable comparisons between consecutive periods. 

Method Used

The seasonal adjustment of Industrial Production Indices is carried out using TRAMO-SEATS methodology. The software that is used for the application of this method is 2.2.2 version of JDemetra+ developed by the National Bank of Belgium (NBB) in cooperation with the Deutsche Bundesbank and Eurostat in accordance with the Guidelines of the European Statistical System.

Seasonal Adjustment Process

Seasonal and calendar adjustment process of Industrial Production Indices begins at the end of each year with the identification of the models of next year.  This specified model structure is kept fixed throughout the year to adjust seasonally and/or calendar effects. At the end of the year, just like the previous year, specification of econometric estimation models for the following year is done. The identified process repeats itself in a cyclical manner each year.

Direct or Indirect Approach

NACE Rev.2 four digit series are seasonally and calendar adjusted by direct approach and then aggregated to derive seasonally and calendar adjusted three digit, two digit figures, MIGs and IPI. The advantages of indirect approach are calculation of contributions to growth from components of main aggregates and retaining additivity.

Revision Policy

The Revision Information Form regarding the revisions to be carried out in the current year for the statistics produced in the scope of this press release can be accessed from the link below.

https://www.tuik.gov.tr/indir/revizyon/2020_RBF_EN_KDIIDB_SUIG_SUE.pdf

Revision Policy

For the Unadjusted Series;

1) When publishing the current month indices previous month indices are revised.
2) The last year indices are revised when the December indices are published.

For the Seasonal and Calendar Adjusted Series;

1) Revisions  are implemented on the data of the last three years excluding the current year.
2) At the end of the year the model structure of the series, model parameters and the outliers observation structure will be changed if the method requires change.

Publication

Adjusted data has been  published in 3 different ways.

1-  “Calendar adjusted” data is derived from unadjusted data by removing calendar and holiday originated effects. Calendar adjusted data should be used in comparisons regarding the same month/period of the previous year.

2-  “Seasonally adjusted” data is derived from unadjusted data by removing effects originating from seasonal effects. Seasonally adjusted data should be used in comparisons regarding the previous month/period.

3 - If unadjusted data contains both calendar and holiday, and seasonal effects, “seasonally and calendar adjusted” data is derived by removing these effects. Seasonally and calendar adjusted data should be used in comparisons regarding the previous month/period.

Not available.