Annual survey on earnings and hours worked

Statistical Office of the Republic of Serbia

Department of Employment and Earnings statistics

Cmilja Ivkovic

Head of Department

5 Milana Rakića, Belgrade

cmilja.ivkovic@stat.gov.rs

+381 11 2412 922 ext. 270

-

The annual statistical Survey on wages and salaries (RAD-1/G) is carried out in order to collect data on salaries and the number of employees depending on their professional qualification, as well as the hours of work and paid bonuses. Data are collected for September each year with an aim to get coefficients which are applied to data from administrative sources (Tax Administration data from the Individual Tax Return Form on calculated taxes and contributions -PPP-PD form), in order to calculate earnings by activities, by degrees of professional education and by gender.

Based on RAD-1/G survey the coefficients necessary for monthly wage estimates by sectors of activity are determined, as well as the coefficients for estimating the average salaries of employees by levels of qualification.

The data on average wages at the level of Serbia - total are based on the data from the PPP-PD form, while data by the NACE Rev. 2 divisions as well as levels of  qualification by gender are estimated based on the structures obtained from the Survey on earnings and working hours.

Average earnings are calculated on the basis of the total amount of accrued salaries and wages for the reporting month and the number of full-time equivalents (FTE) employees.

Classification of Activities (2010), which was adopted by the Government, based on the Law on Classification of Activities (“Official Gazette of RS”, No. 104/09), on the basis of the regulation as of 29 July 2009 (“Official Gazette of RS”, No. 54/10). This classification is without any changes based on Statistical Classification of Economic Activities in EU – NACE Rev. 2.

All sectors of activity from sector A - Agriculture, forestry and fishing to sector S - Other service activities are covered.

The Annual survey on earnings and hours worked (RAD-1/G) covers enterprises that have submitted the financial statements (balance sheets) for the year t-1 (where t is the current year) or they are public funds users in year t, and have 20 or more employees.

Earnings (salaries and wages), in the Survey on wages and salaries (in accordance with the Labour Law and the Law on Personal Income Tax), include all payments to employees to which the corresponding taxes and social security contributions are paid (gross earnings).

In accordance with Article 105 of the Labour Law, earnings of employees with indefinite or fixed-term employment contract include all remuneration for the work done and the time spent at work, earnings based on the employee’s contribution to the business success of the employer (awards, bonuses) and other earnings based on employment, in accordance with the general act and the employment contract.

In accordance with Article 13 of the Law on Personal Income Tax, earnings also include remuneration of employees in temporary or occasional employment (based on contract for performing temporary or occasional jobs – Article 197 of the Labour Law), as well as personal earnings of entrepreneurs.

Earnings of employees include:

• payments for work with full, shorter than full or longer than full-time working hours (overtime),
• allowances for on-call, night and shift work, work on Sundays and public holidays,
• food allowances during work and holiday allowance,
• allowances for hours not worked (annual vacation, paid leave, public holidays, sick leave up to 30 days, absence due to vocational training, a standstill in work not caused by workers’ guilt, compensation because employee is temporarily prevented from working, and compensation is paid by the employer from the first to the last day),
• gifts for children for New Year above the non-taxable amount.

Earnings of employees do not include:

• compensations for sick leave longer than 30 days,
• payments under the service contract or author contract,
• compensation of travel costs to and from work,
• compensation of travel expenses on a business trip in the country or abroad,
• daily allowances for a business trip in the country or abroad,
• accommodation and food expenses, for work and stay in the field if the employer has not provided accommodation and food for the employee,
• retirement or discharge severance pay,
• one-time fees in accordance with the social program,
• compensations for costs of funeral services,
• compensations for occupational injuries or professional diseases,
• jubilee awards,
• solidarity aids,
• participation of the employee in profit,
• compensations of using official cars for official purposes,
• gifts for children up to 15 years of age for Christmas and New Year, to a non-taxable amount,
• premium for voluntary supplementary pension insurance,
• collective insurance premiums against accidents and collective insurance in case of serious illness and surgical interventions,
• reimbursement of expenses for the use of means of work and other labour costs, as well as other remunerations for which the corresponding taxes and contributions are not paid

Average salaries and wages according to the levels of qualification are obtained by dividing the total mass of accrued salaries and wages within the observed degree levels of  qualification by the total number of employees (expressed as the full-time equivalent) working in jobs that require the level of qualification.

Coefficient from 0 to 1.5 is assigned to each employee depending on the number of hours spent at work. Thus, for example, an employee who worked 80 hours in the month in which the full number of working hours is 160 is calculated as 1/2 employee (i.e. the coefficient for the employee is 0.5). In this way, employees who work with several employers and/or in various economic activities and types of ownership, are covered by all employers, and their number is not multiplied, while the earnings are attributed to the activity and the type of ownership in which they are actually realized.

Statistical observation units and reporting units are local units of business entities (companies, enterprises, institutions, cooperatives, and other organizations as well as private entrepreneurs).

For the Аnnual survey on earnings and hours worked, the target population consists of all economically active local units of business entities which according to their legal form are not religious organizations, union organizations, tenants' assemblies or bankruptcy estates.

The sampling frame is constructed using the frozen version of the Statistical Business Register (SBR) from 30th of June of the current year. The sampling frame consists of all local units of business entities (companies, enterprises, institutions, cooperatives, and other organizations as well as private entrepreneurs) which are economically active and have submitted the financial statements for the year t-1 (where t is the current year) or they are public funds users in year t. Sampling frame consists of local units of enterprises with 20 or more employees and local units that according to the SBR have more than 3 employees. The constructed sampling frame included about 33 000 local units of 12 000 business entities.

The Annual Statistical Survey on wages and salaries covers all business entities from the Statistical Business Register (SBR) which have submitted their Individual Tax Application Form on the calculated taxes and contributions through the PPP-PD form.

Republic of Serbia.

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

Until 2018 data on average wages by sex and levels of education were provided on the basis of the Semi-annual survey on employees and their salaries (RAD-1/P).

From 2018, the data on average wages by sex and levels of professional qualification are collected and published annually (for September) based on data from the Tax Administration collected through the Tax Application Form for deduction tax (PPP-PD Form) and data obtained by the Survey on earnings and working hours (RAD-1/G).

    Since September 2019, a new qualification level codebook has been applied to the earnings and hours Survey.

According to the Decision on the Unique Codes for entering and encrypting data in the records by the fields of work (“Official Gazette of RS”, No. 56/18, 101/20) and in accordance with the Law on the National Qualifications Framework of the Republic of Serbia (“Official Gazette of RS”, No. 27/18, 6/20 ), iCode of Degrees, a new system for editing the level and type of qualifications has been established: Register of qualification levels

Not applicable.

Data on paid wages and salaries are collected in Serbian dinars (RSD).

Reference period for the Survey on earnings and working hours (RAD-1/G) is September.

When calculating average earnings, all data from Tax Return forms that are submitted within 45 days after the expiration of the month to which the average earnings relate are included.

Not applicable.

Confidential data are treated in accordance with the prescribed confidentiality principles.

National legislation: Official statistics law.

European legislation: Regulation (EC) No 223/2009 of the European Parliament and of the Council, as of 11 March 2009, on the transmission of data subject to statistical confidentiality to the Statistical Office of the European Communities.

Confidential data are treated according to the 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.

Release calendar can be found on the official SORS website.

Data on average earnings by sex and levels of  qualification, for September of the previous year, are published on the 15^th of April of the current year.

Data on average wages and salaries are published monthly in the statistical release “Average salaries and wages per employee” and press release available on SORS website according to the Release calendar.

Statistical release ZP10 – Average salaries and wages per employee, monthly

Statistical release ZP12 – Salaries and wages of employees by activity, qualification levels and sex, September 

Data on average Salaries and wages of employees by activity, qualification levels and sex are available in the statistical database:

https://data.stat.gov.rs/Home/Result/2403040508?languageCode=en-US

https://data.stat.gov.rs/Home/Result/2403040509?languageCode=en-US

On users’ request, the SORS may provide the data and information obtained by special processing.

-

-

Data are based on reliable data from appropriate administrative sources and surveys. The degree of accuracy is high. Before publishing the results, extensive data quality control is carried out. Compliance with other statistical domains is satisfactory. The release of the results is in accordance with the publication calendar and the data is readily available to all users.

In addition to published data, special processing is done at the request of the user. The most common users are: state authorities, judicial authorities, court experts, financial institutions, economic analysts, media, researchers and students, enterprises and other individual persons.

User satisfaction is measured according to the volume of data downloaded from the official SORS website, as well as by User Satisfaction Survey.

The results of the User satisfaction survey are available on the SORS website.

Not relevant.

Data on average monthly salaries are calculated by combining the estimates of wages and salaries from the statistical survey (RAD-1/G) and data from the Tax Administration (PPP-PD form). Consequently, the results have sampling and non-sampling errors, of which the main are: coverage, nonresponse, measurement, and processing error.

The main sources of errors in this survey (RAD-1/G) are of sampling and non-sampling in nature.

The sample design is such that the estimates for the main domains are of satisfactory precision.

The following non-sampling errors exist in this survey:

1. coverage errors that are due to not completely updated SBR and existing time lag in registering the changes of enterprise characteristics;
2. errors that are a consequence of cut-off, smallest enterprises are excluded from the sampling frame (sampling frame consist of local units that have more than 3 employees of business entities with 20 or more employees);
3. measurement and non-response errors are inevitable, but the process of data collection and data entry are controlled to reduce these errors.

The general assessment is that the type and design of survey and data collection methods ensure sufficient overall accuracy.

The sample design is such that the estimates for the main domains are of satisfactory precision.

The sampling weights are corrected for non-response and the presence of outliers according to the number of employees.

Coefficients of variation of main target variables of the Annual survey on earnings and hours worked are calculated at the level of 2-digit economic activity (NACE Rev. 2. divisions), as well as on the NUTS2 level (region).

On the level of the Republic of Serbia the coefficients of variation for the main target variables: total gross earnings, total net earnings, total taxes, total contributions and number of employees that have received their wages is 0.7%. Coefficient of variation for these variables is lowest in Region of Belgrade (1.5%) and highest in Region of South and East Serbia (2,2%).

The average coefficient of variation of salaries by the level of professional qualification for the Republic of Serbia is 2.5%, and for the wages paid to women is 3.3%.

On the level of Republic of Serbia, the average coefficient of variation for the number of paid employees by the level of professional qualification is 2.5%, and for the number of paid women employees is 3.5%.

The response rate is high, for Serbia in total is 96.3%. The response rate is highest in Region of Vojvodina and add-up to 98.2%, and lowest for the Region of Belgrade and add-up to 93.3%.

Non-sampling errors can be defined as errors arising during all survey activities. The main non-sampling errors in the results on the wages and salaries are coverage, non-response, measurement, and processing errors.

Frame imperfection is the major cause of coverage errors. Auxiliary variables: economic activity, registration status, financial activity and the number of employees are of crucial importance for the definition of the survey sample design and sample selection of the Annual survey on earnings and hours worked. If these variable values are not updated, or with an error, then coverage errors may occur.

Under-coverage errors additionally can arise because the frame lists contain only the data on local units with more than 3 employees and enterprises with 20 or more employees. Furthermore, only enterprises that have submitted the financial statements for the year t-1 (where t is the current year) or are public funds users are covered. Although the magnitude of the under-coverage error is not computed it is not considered to be significant.

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: local unit is closed, extinguished or the complete enterprise is in the process of bankruptcy/liquidation; local unit is economically inactive and local unit has not paid wages for September.

The over-coverage rate for the Republic of Serbia is 3.9%. Over-coverage rate is highest in the Region of South and East Serbia 4.6%.

The main reasons for over-coverage are ‘local unit has not paid wages’ and ‘local unit is closed, extinguished’.

Non-response units are in-scope units for which data were not collected. Non-response units are those for which the response is: runs a business, refuses to submit a report; local unit not found or moved (undeliverable questionnaire) and other reasons (unknown reason of non-response).

The average nonresponse rate was 3.7%. The lowest non-response rate is 1.9% for Region of Vojvodina, and the highest is 6.7% for the Region of Belgrade.

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", which checks whether the questionnaires are filled out completely and does not allow the entry of false data,
• the next step is the release of the program for computation-logical control,
• written and e-mail reminders and telephone contacts with the reporting units are used to keep the non-responses to a minimum, as well as for additional clarifications in case of deviations,
• further steps in data editing are based on the analysis of statistical experts and resulting with correction of discovered errors,
• possible deviations in the data observed during the estimation by the sample methodologist are also checked subsequently.

Data for September of the previous year, are published on the 15^th of April of the current year.

The data are published according to the Release calendar

The same concept of obtaining and calculating average salaries is applied in the territory of the whole Republic of Serbia.

Data are not comparable with the data before 2018.

From 2018, average earnings by municipalities refer to the municipality of residence, while data from Semi-annual survey on employees and their salaries referred to the municipality of work of the employee.

Data from 2018 are not comparable with the data from 2002 to 2017 due to the transit to the use of administrative sources (data from the Tax Administration records).

Not available..

Comparison data from Tax Administration (PPP-PD) does not show major discrepancies between data available from administrative source and statistical survey. A comparative analysis of data on average salaries and wages obtained through the Monthly statistical survey RAD-1 and the data from the Tax Administration records for 2016 and 2017 showed high compliance of data and data compatibility.

Standardized information on cost and burden is not systematically collected.

However, since data from administrative sources (Tax Administration records) are used, thus reducing the survey costs to a minimum, the reporting units are no longer obligated to fill in and submit data every month. Since 2018, data on average wages are calculated by combining the results from the Survey on earnings and working hours and data from the administrative source.

Earlier published earnings data cannot be changed due to the application of average earnings data in many laws and regulations

The source of data for this statistical survey are accounting records and data on the calculated taxes and contributions from the Tax Administration obtained through the PPP-PD form.

The Annual survey on earnings and hours worked (RAD-1/G) is conducted on a two-stage stratified simple random sample. The sampling frame is based on the Statistical Business Register from 30^th of June of current year.

In the first stage, a sample of enterprises is selected. Business entities are stratified according to main economic activity (NACE Rev.2. divisions) and size by number of employees.

Stratification according to economic activity: units of the first stage are classified in 85 classes (in sectors A-S) according to the main economic activity of the enterprise.

Stratification according to the number of employees - enterprises are divided into 3 classes according to the annual average of total employees:

-   with 20 to 50 employees

-   with 50 to 250 employees

-   with 250 and more employees.

Sample allocation of first stage units is according to the assumption that the expected (planned) coefficient of variation of the number of employees (auxiliary variable) on the level of NACE Rev.2. divisions (survey estimation domains) is 1.3%. This way a sample of about 6500 enterprises is allocated. Units of the first stage sample are considered to be clusters out of which local units are selected in the second stage.

The units from the first stage include about 25000 local units. Units in a cluster (business entity) are further stratified according to their main activity (can be different than the activity of enterprise) and number of employees in the local units.

Stratification according to the number of employees (in local units): local units are classified into 2 size groups – bigger and smaller; bigger units are with 100 and more employees.

Stratification according to economic activity: local units (second stage units) are classified in 85 classes according to their main economic activity.

The final stratification of second stage units within the cluster is obtained by cross-classifying size and economic activity.

Sample allocation of second stage units is according to the assumption that the expected coefficient of variation for the number of employees (in local units) on the level of NACE Rev.2. divisions are 4%. The second stage sample size is 11229. The selected second stage sample covered 8611 units belonging to enterprises with 50 or more employees, and 2598 units belonging to business entities with 20 to 50 employees.

Results from the RAD-1/G survey represent a source for determination of coefficients that are applied to the data from PPP-PD forms in order to estimate monthly wages by economic activity, as well as the coefficients for estimating the average annual salaries of employees according to levels of professional qualification and gender.

The frequency of data collection for the Survey on earnings and working hours is annually (September).

Data from Tax Administration. Tax Administration's data are transmitted electronically, through web service, i.e. FTP service (File Transfer Protocol), via a secured network of the Administration of General Affairs of government bodies of the Republic of Serbia. Data are taken on the 15th day of the month.

Data from SBR. Data on economic activity as well as other relevant information are taken from the SBR database.

Survey on earnings and working hours (RAD-1/G). For local units performing activities other than the main activity of the business entity, the data on the masses of paid earnings and the number of employees are assessed on the basis of the structure obtained from the Survey on earnings and working hours (RAD-1/G). For the calculation of data on average wages according to the levels of qualification of employees, the structure of employees and of their salaries from the RAD-1/G survey is applied to the Tax Administration data.

Control of data taken from the Tax Administration is done when processing data on average monthly earnings for September. The structure obtained from the RAD-1/G survey is applied to already controlled data.

Data control is carried out in several phases:

• Contingent control is conducted (comparison of number of records, mass of accrued salaries and wages);
• Logical controls during data entry;
• Control and analysis of data is carried out in order to identify potential errors;
• Comparison and analysis of data from SBR and TA are performed.

Data checks are performed by automated procedures and by experts’ data analysis.

Statistical results (aggregated data) for the reference month are also compared to the results of the previous month and with the corresponding month of the previous year at each level of aggregation.

An imputation method was used to treat non-response. This was done only in exceptional cases (non-response of large units, or those that were specific in their characteristics) when business entities were not able to extract information from the accounting records for the local unit that was selected in the sample. Evaluation of imputation data was based on data on earnings and number of employees according to the level of professional qualifications for the entire business entity and with the help of information from the SPR on the number of employees in local units. Information about unit non-response is taken into account for adjustment of weights.

The estimations of totals are calculated according to the standard procedure for the two-stage stratified random sample (Horvitz-Thompson estimates). This means that estimates of totals calculated by summing the weighted values from the sample. Final weights are a multiplication of the weight from the first stage and weight from the second stage.

First stage weight for the units (enterprises) of the same stratum is equal to the quotient of the number of units in the stratum (planned sample) and the number of units in the realized sample in the stratum.

Second stage weight is the weight of local units in a stratified cluster (business entities), weight for the units of the same stratum is equal to the quotient of the number of units in the stratum and the number of units in the realized sample in the stratum.

The realized sample consists of:

• the units that filled out the questionnaire,
• units for which data is evaluated and
• units that are a part of the over-covered units for which the response information is one of the following: local unit is economically inactive and local unit has not paid wages for September.

The rest of the units are classified into the unrealized sample. An unrealized sample also includes part of over-coverage units (for reasons: local unit is closed, extinguished or the complete enterprise is in the process of bankruptcy/liquidation) in order to compensate, by number, newly created units that a sampling frame does not contain.

Local units detected as outliers with respect to the number of paid employees are put into special census strata. The weights of the remaining units in the strata from which they originated were corrected.

-