Statistical Office of the Republic of Serbia

Two organisational units take part in the activities on the variable Turnover in services. All relevant information relating to the Catering activities are marked with A), and for other sections covered by the mentioned indicator with B).
Information that is in common for both organisational units are not be specially marked.
A) Catering and tourism statistics group
B) Structural business statistics section

Republic of Serbia, Belgrade, Milana Rakica 5

The turnover indices for services are important indicators for monitoring the dynamics of movements in this part of non-financial economy as well as for short term forecasts. Services turnover indices are derived based on the data on turnover from three statistical surveys and administrative sources (Tax Administrative Data Base). The statistical surveys are:

A) Quarterly Surveys on Catering Trade (UG-11, UG-11P forms) provide the data on the movements, volume and structure of turnover generated in the catering trade activity by sale of catering services (accommodation, food, beverages) and other services usually rendered in catering trade.
In order to achieve these goals, the following data are collected by surveys:
• turnover of business subjects for the reference quarter and by months;
• turnover generated by catering activities by statistical territorial units (NSTJ 2);
• the structure of turnover generated from catering activity by types of catering services.

B) Quarterly Structural Business Survey (SBS-03 form) provides data on financial operations of enterprises which is based on: operating income, operating costs and stocks. Within "Operating income", data for "Revenues from sales of goods, products and services", which in fact represent generated turnover.
Since 2017, the questionnaire has been extended with questions on monthly revenues from sales of goods, products and services for the reporting quarter. The monthly estimates of turnover are calculated by domains that comply with the European Business Statistics Regulation (EBS).
These surveys refer to enterprises (business subjects) that in the legal world correspond to legal persons, with  the exception of the Quarterly Survey on Catering Trades (form UG-11P), which refers only to unincorporated enterprises. For unincorporated enterprises in other divisions of the Classification of Activities (2010), data are extracted from the Tax database from the VAT return.
The indices are calculated by combining turnover estimates from these three statistical surveys and data from the Tax Administration Database (form PPPDV). 
The calculation of monthly turnover indices in service activities is the result of the implementation of subproject 3.2. Short-term business statistics within the Statistical Cooperation Program with the European Union IPA 2019 Multi-beneficiary Statistical Cooperation Programme. The program was implemented in the period from September 2021 to March 2024.

The National Classification of Activities - CA 2010, that completely corresponds to the Statistical classification of economic activities in the European Community (NACE Rev.2):

http://www.stat.gov.rs/media/2319/zakon-o-klasifikaciji-delatnosti.doc

The surveys listed under item 3.1 have the following coverage:
A) Quarterly Surveys on Catering Trade (UG-11, UG-11P forms):
I Accommodation and food service activities
I 55 – Accommodation
I 56 – Food and beverage service activities
These surveys cover enterprises and unincorporated enterprises whose principal activity is classified in Section I (NACE Rev. 2).
B) Quarterly Structural Business Survey (SBS-03 form):
This survey covers enterprises whose principal activity is classified into non-financial business economy (Sections A-S, excluding K and O). For the purpose of production of services turnover indices, the following sections and divisions of NACE Rev. 2 (except I) are covered.
H 49 – Land transport and transport via pipelines
H 50 – Water transport
H 51 – Air transport
H 52 – Warehousing and support activities for transportation
H 53 – Postal and courier activities
J 58 – Publishing activities
J 59 – Motion picture, video and television programme production, sound recording and music publishing activities
J 60 – Programming and broadcasting activities
J 61 – Telecommunications
J 62 – Computer programming, consultancy and related activities
J 63 – Information service activities
L 68 – Real estate activities
M 69 – Legal and accounting services
M 702 – Management consulting services
M 71 – Architectural and engineering activities; technical testing and analysis
M 73 – Advertising and market research
M 74 – Other professional, scientific and technical activities
N 77 – Rental and leasing activities
N 78 – Employment activities
N 79 – Travel agency, tour operator reservation service and related activities
N 80 – Security and investigation activities
N 81 – Services to buildings and landscape activities
N 82 – Office administrative, office support and other business support activities

Level of data processing: divisions-level (2-digit), European level for dissemination M 70.2 are estimated on the basis of the turnover estimates on the level of M 70 and the corresponding structures of turnover from sampling frame (auxiliary variable).
At higher level of aggregation, turnover indices have been prepared for:
Services – total;
H – Transportation and storage;
I – Accommodation and food service activities;
J – Information and communication;
L – Real estate activities;
M – Professional, scientific and technical activities, and
N – Administrative and support service activities.
Administrative data source - VAT returns obtained from the Tax Administration are used in order to provide data on turnover for unincorporated enterprises from all covered sections excluding Section I.
There is no threshold by size of businesses (according to the number of employees) - all size classes are covered by the surveys and administrative data source.

The objective of the turnover index is to show the development of the market for goods and services.

Turnover comprises the totals invoiced by the observation unit during the reference period, and this corresponds to market sales of goods or services supplied to third parties.  Turnover includes all costs charged to the customers (transportation, packaging, insurance, etc.) even when those costs are shown separately in the invoice. Turnover excludes rebate, discount, cash accounts, excise, VAT and other taxes which are calculated at the time of products sales.

The statistical unit is kind of activity unit (KAU).
The reporting unit is an enterprise.
Administrative data source - VAT returns obtained from the Tax Administration used in order to provide data on turnover generated by unincorporated enterprises and annual financial statements.

The target population consists of enterprises with main economic activity is one of the NACE activities listed under item 3.3.
The sampling frames for statistical surveys are constructed using the frozen version of Statistical Business Register (SBR) from 31st of December of the year t-1, where t is the current year. The sampling frames were constructed from these lists of enterprises by excluding the smallest enterprises. Annual turnover and size of enterprises are taken from financial statements for the year t-2, where t is the current year.

The surveys listed under item 3.1 use different criteria to exclude the smallest units from the sampling frames with respect to turnover from the financial statements:

A) I: Enterprise is excluded if its turnover is less than or equal to 2500 thousand RSD; 5% of the smallest unincorporated enterprises according to turnover (by NACE divisions) were excluded;

B) In other NACE Rev.2 Services activities: The percentage of excluded turnover is less than or equal to 4%, by NACE divisions. 

The tax Administration Database (monthly and quarterly) is an additional source on turnover for unincorporated enterprises which are not covered by statistical surveys. Turnover data of about 8.6 thousand units from Tax administration is added to the survey estimates of turnover.
For statistical surveys, the number of enterprises/ unincorporated enterprises in the SBR population and the sampling frame in Services NACE Rev. 2 activities are presented in the Table below.

Table 1. Number of units in statistical surveys for Services in 2023

The reference area is the Republic of Serbia.

Remark: Starting from 1999 the Statistical Office of the Republic of Serbia has not at disposal and may not provide available certain data relative to AP Kosovo and Metohia and therefore these data are not included in the coverage for the Republic of Serbia (total).

The turnover of the units located outside the Serbian territory is not included.

For all the series of the index, the data are available from January 2017 and onwards.

Base year: 2021 = 100. The same base and reference periods are published at the national level, starting from March 2024 (reference month: January 2024)

Indices (2021 = 100) are unadjusted, calendar adjusted and seasonally adjusted.

The reference period is a calendar month.

The statistical surveys listed under item 3.1 are carried out on the basis of the following legal acts of the Republic of Serbia: 

The official statistics law (Official Gazette RS, No 104/09): 

http://www.stat.gov.rs/media/2322/zakon_o_statisticie.pdf

Decision of  the Programme of Official Statistics over the period 2021 – 2025 (Official Gazette RS, No 24/21)

https://www.stat.gov.rs/media/5948/program2021-2025_engleski.docx

Regulation on the Plan of official statistics for 2023 (Official Gazette of the RS, No. 141/22) 

https://www.stat.gov.rs/media/377224/planzvanicnestatistike2023_engleski.pdf

Law on the Classification of activities (Official Gazette RS, No 104/09):

http://www.stat.gov.rs/media/2319/zakon-o-klasifikaciji-delatnosti.doc

Relevant legal and other documents on European level are:

European Business Statistics (EBS) Regulation (EU) 2019/2152 of the European Parliament and of the Council and Commission Implementing Regulation 2020/1197 laying down technical specifications and arrangements pursuant to the mentioned EBS Regulation (General Implementing Act).

Non-confidential data are available on the SORS website. Data on services turnover are submitted to Eurostat via eDamis, in regular monthly periodicity, in the form of unadjusted, calendar-adjusted and seasonally adjusted indices. 

At the national level data confidentiality is stipulated by:

Official Statistics Law (Official Gazette of RS, number 104/2009), Articles 44, 45, 46, 47, 48, and 49.

http://www.stat.gov.rs/media/2322/zakon_o_statisticie.pdf

Rulebook on statistical data protection in the Statistical Office of the Republic of Serbia

http://www.stat.gov.rs/media/2343/rulebook-on-statistical-data-protection-in-sors.doc

Guidelines on measures of data and information protection in the Statistical Office of the Republic of Serbia

http://www.stat.gov.rs/media/2341/guidelines-on-measures-of-data-and-information-protection-in-the-sors.doc

European level:

Regulation (EC) No 223/2009 of the European Parliament and of the Council of 11 March 2009 on European statistics.

Confidential data are treated according to Official Statistics Law and Rulebook on Statistical Data Protection in the SORS of the RS, article 3. "Confidential statistical data, in terms of this Rulebook, are individual data as well as aggregates based on which the data holder may be in whatever way recognized", article 11 "Confidential data from Article 3 of this Rulebook are considered official secret and cannot be published or communicated, that is, they cannot be part of aggregated data from which individual data can be identified".

Data are published on the level at which they are not confidential. 

A) Calendar of releases is available at SORS website on the 1st day of December every year for the forthcoming year: 

http://www.stat.gov.rs/en-US/calendar. 

B) Not available.

A) Calendar of releases is available at SORS website: http://www.stat.gov.rs/en-us/calendar/

B) Not available.

The data obtained through statistical surveys are released at the:

- internet presentation: http://www.stat.gov.rs/en-US/,

- in the database: http://data.stat.gov.rs/?caller=SDDB&languageCode=en-US,

- in statistical publications: http://www.stat.gov.rs/en-US/publikacije/.

The results of official statistics are available at the same time to all users under the same conditions, and any privileged access by external users before their release is not allowed.

The frequency of dissemination is monthly.

Not available.

A) Data on Section I – Accommodation and food service activities are published in the Monthly Statistical Bulletin, Statistical Yearbook of the Republic of Serbia and Statistical Pocketbook of the Republic of Serbia.

B) Not available.

A) Data on Section I – Accommodation and food service activities is available in the Statistical Database on

http://data.stat.gov.rs/Home/Result/220101?languageCode=en-US&displayMode=table and

http://data.stat.gov.rs/Home/Result/220102?languageCode=en-US&displayMode=table

B) Monthly indices of turnover in services in the Republic of Serbia are available on SORS Statistical database on 

https://data.stat.gov.rs/Home/Result/290104?languageCode=en-US  

On request, the access to anonymized microdata is ensured to scientific and research institutions. The request can be placed by email at stat@stat.gov.rs, or by regular mail addressed at 5 Milana Rakića St, Belgrade.

For more information, please see http://www.stat.gov.rs/en-US/korisnicka-podrska/micropodaci.

Monthly indices of services turnover (unadjusted, calendar-adjusted and seasonally adjusted), starting from January 2017, base year 2021=100, are submitted to Eurostat in accordance with the required level of detail, within the specified deadlines and in the required format. 

A)  Questionnaires for Accommodation and food services turnover are available on SORS website:

http://www.stat.gov.rs/istrazivanja/methodology-and-documents/?a=22&s=0

Methodological explanations (metadata) are available at SORS website, option "M":

http://data.stat.gov.rs/Home/Result/220102?languageCode=en-US

B) Content of the Quarterly structural business survey form is available on the SORS website:

http://www.stat.gov.rs/en-US/istrazivanja/methodology-and-documents/?a=19&s=1904

Methodological explanations for turnover in services are available within on SORS Statistical database, option "M":

https://data.stat.gov.rs/Home/Result/290104?languageCode=en-US

Documentation on quality is not yet available.

The SORS quality management system is relied on the Serbian official statistics mission and vision, as well as on the European Statistics Code of Practice – CoP and the Total Quality Management – TQM principles, which together make the common quality framework of the European Statistical System (ESS).

For more information, please see the documents at: http://www.stat.gov.rs/en-US/o-nama/sistem-upravljanja-kvalitetom.

Through all phases of data production, main quality aspects are controlled by statisticians, sample methodologist and IT experts. 

The main users of these data include State authorities, financial institutions, business subjects, economic analysts, the media, researchers, students and other individual users.
On international level, the Statistical Office of the European Union (Eurostat) is the user of data.
Within the statistical system, the main user of these data is the Department of National Accounts.

On biennial basis (once in two years) the User Satisfaction Survey is implemented through a web questionnaire. 

The survey results are available at SORS website: http://www.stat.gov.rs/en-US/o-nama/sistem-upravljanja-kvalitetom.

The User Satisfaction Survey that refers particularly to short-term indicators has not yet been conducted.

A) Indicators are harmonized with the concepts and definitions in the methodology based on the EU Council Regulation (EU) 2019/2152 of the European Parliament and of the Council on European Business Statistics. Comparable series of Turnover Services Indices for Section I (Accommodation and food service activities), Divisions I 55 (Accommodation) and I 56 (Food and beverage service activities), could be recalculated from the January 2008 onwards, but for the consistency of the whole series that includes all service activities the indices has been calculated since January 2017.

B) Indicators are harmonized with the concepts and definitions in methodology based on the EU Council Regulation (EU) 2019/2152 of the European Parliament and of the Council on European Business Statistics. The requirements on the length of the time series are not fulfilled. The reasons are as follows:

Quarterly Structural Business Survey (SBS03) has introduced the collection of monthly data (in the reference quarter) on the turnover in activities in-scope of the ‘Services’ requirements since 2017 for the enterprise as the statistical unit, although data collection still remains to be quarterly. In Serbia, the implementation of the EBS Regulation started in January 2023. Then the indices were calculated with a base of 2015=100 and published in the SORS dissemination database and submitted to Eurostat. In order to provide data both on KAU level and in monthly periodicity, the following was done:
1. Recalculation of data from enterprise level to KAU level
Developing an input-output matrix as the relationship between turnover of the main economic activities and all the kind-of-activities using the micro-data from the Annual Structural Survey on Business Entities (SBS01A form) of the 300 largest enterprises in non-financial business economy.
The total turnover of the 300 largest enterprises was decomposed by determining the share of the enterprise's total turnover coming from activities other than the main economic activity of the enterprise. Then, based on the decomposition of the largest units an input-output matrix was formed by establishing the total share of turnover in each NACE division originating as revenues from other NACE activities, in other words as revenues from Kind-of Activity Units (with activity different of the main economic activity) of those largest enterprises.
Detailed analysis showed that this set of the 300 largest enterprises is sufficient to represent the economic trend of the total in the non-financial business economy, the input-output matrix was used to calculate the necessary increase or decrease of the total turnover on the enterprise level when compared to the totals on the KAU level. Those, giving the requested totals of turnover on the Kind-of-Activity level.
Data from the Annual Structural Survey on Business Entities is available from the years 2018 and onward. The necessary turnover totals on KAU level for 2017 were calculated by using the input-output matrix corresponding to the year 2018 as a proxy. Data from this survey are internally available nine months after the calendar year. Therefore, micro-data concerning the year 2023 will be available in September 2024. Because of this, for the purpose of calculating the needed results on the KAU level for the reference year of 2023, the input-output matrix corresponding to the year 2022 was used as a proxy. Also, the same matrix will be used for calculations in the current year until more up-to-date data won’t be available (when a revision of results will be conducted).

In 2022, there were significant changes in the structure of the main activity of several large reporting units. Those units, due to their size, have a significant impact on the entire economy, and their parts (KAU) significantly affect those sectors and areas of KD 2010 that had a small share in the structure of total turnover. At the level of the total turnover of services, according to ENT and KAU, an average growth of turnover of 4% is noted. Therefore, a revision of the value of turnover at the level of activity type units (KAU) was carried out for the entire series and the indices were recalculated.
2. Recalculation of data from quarterly to monthly time series
Since monthly data on turnover in service activities for companies have been collected (quarterly) since 2017, the above described method for calculating totals of turnover on the KAU level was possible for monthly turnover totals for 2017 and onwards.

In order to calculate the indices based on 2015=100, it was necessary to estimate the monthly turnover for 2015 and 2016. This was done by "temporal disaggregation" of quarterly time series, at the KAU level obtained by applying the method described above and using monthly turnover data from the Tax Administration database as an auxiliary time series.

Since the applied method for temporal disaggregation is based on regression models, producing longer high-frequency time series reduces the quality of the estimation. Therefore, it was decided that the starting point of the revised monthly time series with a base of 2021=100 will be January 2017, when SORS has monthly data collected from the SBS03 survey.

Data on turnover in services are collected in statistical surveys or are derived on the basis VAT returns, obtained from the Tax Administration. Consequently, the results have sampling and non-sampling errors, of which the main are: coverage, nonresponse, measurement, and processing error.

The design and organisation of the surveys, as well as the data collection and processing methods ensure sufficient overall accuracy.

Tax administration data is available 30 days after the reference month (quarter). For units that report VAT data on a monthly basis, quarterly values of turnover are obtained by summing the monthly values. The quarterly turnover estimates in services are calculated by summing units’ quarterly turnover values; for units that report VAT data on a quarterly basis, monthly values for the three months of the quarter are estimated.

The sampling designs of the statistical surveys ensure results of satisfactory precision for the estimates of totals required by Eurostat.
The sampling weights are corrected for non-response and for the presence of outliers.

Coefficients of variation are calculated for estimates of turnover totals by NACE Rev. 2 divisions and groups that are in the coverage of Services Turnover, as well as for corresponding sections and the Republic of Serbia. On the level of the Republic of Serbia, for 2023 coefficients of variation for monthly turnover are around 2.1%. Monthly average оf the coefficients of variation is highest in divisions N 74 (19.1%), J 59 (17.1%), N 79 (15.7%), M 70.2 (15.3%) and J 58 (13.2%). Monthly average of the coefficients of variation is lowest in divisions H 53 (1.2%), H 51 (1.4%), and J 62 (4%).

The response rate is high in all quarters. On the level of the Republic of Serbia, the rate is between 82.2% (first quarter) and 86.2% (second quarter).

The main non-sampling errors in the results on the turnover in services are coverage, non-response, measurement, and processing errors.
Frame imperfection is the major cause of coverage errors. SBR variables: an economic activity of the enterprise, registration status and turnover from the financial statement for the year t-2 are of crucial importance for the definition of the survey's sample designs and sample selection. If these variable values are not updated, or with an error, then coverage errors may occur.
In addition, the under-coverage errors arise because the frame lists contain only the data on enterprises that have submitted the financial statements for the year t-2. Those that opened after that year are not covered unless it is known that they are exceptionally big or important for the activity they perform. Under-coverage errors are caused also by exclusion the smallest enterprises from the sampling frame with respect to turnover (from financial statements for the year t-2).
Although the under-coverage error is not measured, it is assumed that it does not significantly affect the results because the survey does not include primarily small enterprises.
Over-coverage and non-response errors (or response rates) are calculated using information about the response of the unit.
The units of over-coverage are those for which the response is: closed, extinguished; blocked, not working; outside the scope of the survey (does not perform the selected activity); bankruptcy and liquidation. 

Based on the results for 2023, the over-coverage rate for the Republic of Serbia is in the range of 2.8% (first quarter) to 6.4% (fourth quarter).

Units are classified as non-response if the response is: runs a business, but refuses to submit a report; not found or moved (undeliverable questionnaire) and other reasons (unknown reason of non-response). On the opposite, all other in-scope units are classified as the response. These are the units that, at the data collection phase, have provided data.

Based on the results for 2023, the response rate for the Republic of Serbia is in the range of 82.2% in the first quarter to 86.2% in the second quarter. The size-weighted response rate is in the range of 85.6% in the fourth quarter to 88.6% in the second quarters. As a measure of size the turnover from sampling frame (auxiliary variable) is used.

In order to reduce the non-response, the following measures are undertaken: web questionnaire is introduced, reminders are sent in written or e-mail form, reporting units are contacted by telephone. Once a year, a new coordinated sample with rotation is selected (using the SORS system of coordination). In this way, a gradual replacement of small sampling units is provided, among which non-response is frequent. Every effort is made to reduce non-sampling errors to a minimum.
In order to reduce non-sampling errors caused by human errors, the following measures are undertaken: collected data are entered in the PC application that represents the so-called "controlled entry". During this phase, the program does not allow the entry of numerically inconsistent data. The next step is to run the program for checking logical inconsistencies. Further steps in data editing are based on the analysis of statistical experts and resulting with correction of discovered errors. If necessary, the reporting unit is contacted in order to get the correct data. Finally, data validation is realized during analysis of the aggregates.

In the survey SBS-03, data imputation is carried out for non-responding units whose non-response could significantly increase the bias of the results. These are, above all, extremely large units in their field of activity or those that are specific in some of their characteristics. The imputation method is constantly evolving, and for auxiliary variables, various sources of data are used such as tax information, final accounts, and survey data.

Based on preliminary results for SBS-03 for the 2023 and referring only to Services excluding divisions I 55  and I 56, the imputation rate for the Republic of Serbia is in the range of 0.48% (second quarter) to 0.74% (fourth quarter).

The first results of surveys are available 60 days after the end of the reference month. 

A) There is no time lag between the actual delivery of the data and the target date when it should have been delivered. Deadlines are respected and data are published and delivered on time, according to the pre-announced publications calendar.

B) Turnover indices in services are published in the dissemination database 60 days after the end of the reporting month. The data are transmitted to Eurostat in the required deadlines and technical format.

Turnover indices in services are representative only for level of Serbia – total. They are not representative for the lower regional levels.

All data series of the presented indicators are comparable in time.

The data obtained by all three statistical surveys (Quarterly Surveys on Catering Trade and Quarterly Structural Business Survey) are used as an input data for the calculation of the quarterly macroeconomic aggregates. It means that the degree of their compliance is the highest possible.

The data from quarterly surveys and annual SBS are not completely coherent. The reasons are the different purposes and hence different used methods for the statistics. There are also some differences in the European definitions. 

The data from statistical survey are checked for internal coherence. Each level of aggregated data is compiled directly from micro data. Control is also done by comparing the value of turnover from Quarterly Structural Business Survey with corresponding data from the Department of Statistics of Trade, Catering and Tourism; with annual financial statements and with monthly and quarterly data from Tax Administration.

Standardised information on cost and burden is not systematically collected.

A system of coordination of random samples of business systems has been introduced, which reduces the burden of small units.

More details in publication "Sample coordination of statistical business surveys":

http://publikacije.stat.gov.rs/G2014/PdfE/G201410089.pdf

The SORS general revision policy constitutes the global frame ensuring that each statistical domain shall define its own revision policy in compliance with its specific nature.

The general SORS revision policy determines:

- general rules of revisions of the published data,

- forms of informing users as regards the possible causes of revisions,

- categories of revisions, and

- documents covering all aspects of revisions.

The general SORS revision policy is available at SORS website:

http://www.stat.gov.rs/media/2332/general-revision-policy.docx

The same revision policy is applied for publishing of national data and for data transmitted to Eurostat.

So far, thе national calendar revision does not exist. 

The benchmarking of turnover data from the Quarterly Structural Survey and data from the Quarterly Survey of Catering is regularly conducted. Also, comparisons are made with monthly and quarterly data from the Tax Administration database, as well as with annual financial reports.

A) All published monthly turnover indices for Section I of NACE Rev. 2 are treated as preliminary, which means that certain corrections are possible based on subsequently obtained data from quarterly surveys (by month), which are treated as final. If errors are detected, corrected data are published and disseminated as soon as possible. Major revisions are due to changes in the methodology, classification of activities (e.g. the implementation of NACE Rev. 2) and base year change. Routine revisions of Accommodation and food services (Section I of NACE Rev. 2) data are due to new or corrected data received from reporting units or administrative sources. The corrected data are published together with the previous data in the following quarter with an appropriate explanation. All published monthly indices are treated as the previous, which means that some corrections are possible based on subsequently obtained information.

B) Two revisions are made for turnover indices in service activities (except for sector I) for all months of the current year:

1) Estimated and preliminary data: 

Given that monthly data are collected quarterly, for the first two months of the reporting quarter turnover data are provided by means of forecast estimation. All forecast estimations for the first two months are made in the JDemetra+ 2.2.4 software, using the X13 ARIMA - SEATS method. By the time when the data have to be transmitted for the third month of the quarter, the quarterly data collection and sample estimations are completed. Thus, at this point the estimation for all three months of the reporting quarter are calculated as survey estimates of turnover from the SBS03 survey (sample estimation of enterprises for total revenues from sales, and turnover from the VAT for unincorporated enterprises and recalculated onto KAU level), and that is considered as preliminary data;

2) Final data:

Estimates are recalculated 10 months after the reference year, when additional data is available (such as annual financial statements, the complete Tax Administration Data Base). These data are used to check, benchmark and detect errors in survey data. Sample estimation with corrected data and with included data of enterprises that were received after the quarter deadline are considered as final and are available to the users in the SORS database the last day of October next year: 

https://data.stat.gov.rs/Home/Result/290104?languageCode=en-US

In order to fulfill the requirements of the EBS regulation turnover, data from enterprise level is recalculated to turnover on KAU level. Microdata of the Annual Structural Survey on Business Entities (SBS01A) for the 300 largest enterprises in non-financial business economy are used for the recalculation. The total turnover of the 300 largest enterprises was decomposed by determining the share of the enterprise's total turnover coming from activities other than the main economic activity of the enterprise. Then, based on the decomposition of the largest units an input-output matrix was formed by establishing the total share of turnover in each NACE division originating as revenues from other NACE activities, in other words as revenues from Kind-of Activity Units (with activity different of the main economic activity) of those largest enterprises. The established input-output matrix was used to calculate the necessary coefficient of increase or decrease of the total turnover on the enterprise level when compared to the totals on the KAU level. Data from the Annual Structural Survey are internally available nine months after the reference year. Therefore, micro-data concerning the year 2023 will be available in September 2024, when the appropriate input-output matrix will be developed for the reference year 2023 and it will be used to revise the estimated monthly turnover on KAU level for 2023 and each subsequent year and transmitted monthly data up to that point.

As already described in chapter 12.3, due to significant changes in the structure of the main activity of large reporting units in 2022 and their impact on the calculation of turnover according to KAU, a significant revision of the turnover for the entire series was carried out and the indices were recalculated.

Routine revisions are promptly implemented. The type of revision is clearly highlighted in the published tables.

Routine revisions, apart from other objectives (such as improving accuracy and reliability, etc.) aim to achieve temporal consistency.

In 2024, the planned change of the base year was carried out and all indices of turnover in services were rebasing from the base year 2015 (2015=100) to the base year 2021 (2021=100).

The series starts with January 2017 because that is the first period with reliable monthly data on Services Turnover from statistical surveys available in SORS.

All aggregates and components are revised at the same time. They have been analysed in detail and have no breaks in the time series. Basic information about revisions can be found in the methodological explanations and are available to all users at the same time.

So far, the national calendar revision does not exist.

The data obtained by all three quarterly surveys are used as an input data for the calculation of the quarterly macroeconomic aggregates. It means that benchmarking is carried out within the Units of Quarterly National Accounts.
The data from quarterly surveys and annual SBS are not completely coherent. The reasons are the different purposes and hence different used methods for the statistics. There are also some differences in the European definitions. However, the Annual Structural Statistics Unit carried out benchmarking at microdata level with Quarterly Structural Business Survey, nine months after the reference year.
Mean Revision (MR) and Mean Absolute Revision (MAR) are used as revision indicators.
Given that the monthly data series are being calculated for the first time, the MR and MAR indicators have not yet been calculated. 

The source of data for calculation of Turnover in Services, are statistical surveys or administrative data (VAT returns obtained from the Tax Administration).

Surveys are based on random samples. The sampling design of these surveys is stratified random sample. Specific characteristics of the samples are:

A) Quarterly Survey on Catering Trade for Enterprises (UG-11):
The sampling frame units are stratified according to NACE Rev. 2 divisions, NSTJ 2 (National NUTS 2 territorial classification) and an indicator of census strata. Census strata are strata that contain units that have significant turnover according financial statement.
    Quarterly Survey on Catering Trades for Unincorporated enterprises (UG-11P):
The stratification of the sampling frame is according to to NACE Rev. 2 divisions, indicator of TAX payer, size and NSTJ 2 territorial classification. Four size classes are determined with respect to the values of turnover from financial reports. The largest units form the size class whose units are completely enumerated. Other three size classes are of equal size.

B) Quarterly Structural Business Survey:
The sampling frame units are stratified according to NACE Rev. 2 divisions, two classes according to the number of employees (less than 50 employees and 50 and more) and two classes according to turnover (census, and non-census). Census strata are strata that contain the largest units, with respect to turnover.
The auxiliary variable values on turnover are from financial statements, while the number of employees is a twelve month average on the number of employees who have an employment contract (definite or indefinite). The source for employees is The Central Register of Compulsory Social Insurance (CRCSI), data that is available at the time of the construction of the frozen version of SPR of December of the year t-1.
Remark: To fulfil national requirements, a more detailed stratification of the frame was defined so that the activities in the coverage of the Information and Communication Sector are planned domains. In these domains, higher precision was set for the estimates of auxiliary variables totals.

In all surveys, sample allocation of units is made according to Bethel algorithm, which is based on a Neyman`s allocation method. The overall sample size and the allocation among the different strata is determined starting from the constraints imposed in the survey on the accuracy (i.e. planned errors in terms of coefficients of variation) of the estimates of interest or in this case on auxiliary variables (turnover/number of employees) by main survey estimation domains.
Samples are selected using sequential sampling scheme based on permanent random numbers from the uniform distribution on (0,1). This method of selection, with the rotation of sampling units, is a part of SORS system of coordination that is based on the SAMU system (SAMU, 2003). See Sample Coordination оf Statistical Business Surveys.

Table below includes data on sample sizes, sample fractions and net samples by surveys.

Table 2. Sample size, sample fraction and net sample size for the year 2023

Note: For calculation of final results for Other Services (excluding I), the net sample size will be larger.

In statistical surveys, data are collected quarterly. All forms collect data by month of the reporting quarter.
Tax administration data are at disposal: monthly (for large taxpayers) about 30 calendar days after the end of the month and quarterly (for small taxpayers) 30 days after the end of quarter.

In statistical surveys, the method of collection is by post, by e-mails or by Web application. Questionnaires are sent to reporting units before the expiration of the referent quarter. Questionnaires are also available on the SORS Website. Alternatively, questionnaires are collected using fax machines.

For unincorporated enterprises, data from VAT declarations are used. Data is downloaded via FTP (File Transfer Protocol) server.

Actions taken to speed up or increase the rate of response are urgency letters by post, telephone contacts, and e-mails.

For all surveys, data validation is carried out through the chain of activities: control of collected questionnaires by statisticians in regional offices; data entry into the PC by an interactive application which does not permit some crucial errors; detailed checking and correction of data in the program for batch editing; data editing during the analysis of data and the estimation procedure; data validation during analysis of the aggregates. If necessary, reporting units are contacted in order to get correct data.

The methods used to treat non-response: imputation for large non-respondents or those with some specific characteristics (Services activities other than I 55 and I 56); the other units are treated by reweighting (weights are adjusted to take account of non-response).
The standard estimation procedure for the stratified simple random sample is used. Estimates of totals are calculated by summing the weighted sampling values. The sampling weight is equal to the ratio of the number of units in the stratum to the number of stratum units in the sample. It is adjusted for non-response. Also, enterprises that are marked as outliers are extracted into special census strata. The weights for the remaining units of a stratum from which an outlier originated are corrected.
The service indices are compiled as simple indices (value relative). The indices are calculated directly from the data.

More information about metadata calendar and seasonal adjustment is in the Annex.

No comment.