Reference metadata describe statistical concepts and methodologies used for the collection and generation of data. They provide information on data quality and, since they are strongly content-oriented, assist users in interpreting the data. Reference metadata, unlike structural metadata, can be decoupled from the data.
National Statistics Office, Lascaris Valletta, VLT 2000, Malta
1.6. Contact email address
Confidential because of GDPR
1.7. Contact phone number
Confidential because of GDPR
1.8. Contact fax number
Confidential because of GDPR
2.1. Metadata last certified
13 June 2025
2.2. Metadata last posted
13 June 2025
2.3. Metadata last update
13 June 2025
3.1. Data description
The short-term statistics labour variables are the number of employees and self employed persons, hours worked by employees and gross wages and salaries.
These labour variables are compiled for the industry, construction, wholesale and retail trade and services domains.
However, in practice, the self employed persons sub-variable is not observed.
3.2. Classification system
The classification used in STS is NACE rev.2.
3.3. Coverage - sector
Industry domain: NACE Sections B, C, D and Division E36.
Construction domain: NACE Section F.
Wholesale and retail trade domain: NACE Section G [NACE Divisions 45, 46 and 47].
Services domain: NACE Sections H, I, J, L, M, N.
3.4. Statistical concepts and definitions
The definitions of the main variables for the domains of industry, construction, trade and services are the following:
Employment: The number of employees includes all paid and unpaid workers. Self-employed persons are also included.
Hours worked by employees: Hours worked includes the total amount of hours actually worked, that is: regular working hours and also during overtime; during nights, Sundays or public holidays. Hours worked but not paid are also considered. Hours worked does not include vacation and sick leave, loss of work-hours due to accidents, strikes, lock outs, slack time etc. and also time spent for meal breaks and commuting strike days. The hours worked for self-employed persons are not included.
Gross wages and salaries:Gross wages and salaries are defined as the total sum of remuneration in cash or in kind payable to all persons employed in return of work done during the reference period, irrespective of whether this remuneration is paid regularly or not. Direct taxes and employees’ social security and pension contributions are included. Wages and salaries of self-employed persons are not included.
3.5. Statistical unit
The statistical unit is the Kind of Activity Unit (KAU).
3.6. Statistical population
Industry variables: The statistical population is around 2000 units and includes all those units falling under NACE divisions 8 - 36.
Construction variables: The statistical population is around 3300 units and includes all those units falling under NACE divisions 41 - 43.
Wholesale and retail trade variables: The statistical population is around 9000 units and includes all those units falling under NACE divisions 45 - 47.
Services indicators variables: The statistical population is around 6000 units and includes all those units falling under NACE sections H, I, J, L, M and N.
3.7. Reference area
The indicators cover the national territory of the Maltese Islands. They do not cover activities outside this geographical area.
3.8. Coverage - Time
The data for these indicators are available from the year 2000.
3.9. Base period
Base year (year on which weights are based): 2021
Reference year (year when the index equals 100): 2021
The same base and reference years are used for data published locally and that transmitted to Eurostat.
The data are provided in index form both to Eurostat and when publishing news releases on a national level. Percentages changes from one period to another are also published.
Industrial labour variables are submitted on a monthly basis.
Construction labour variables are submitted on a quarterly basis.
Retail trade labour variables are submitted on a quarterly basis.
Services labour variables are submitted on a quarterly basis.
6.1. Institutional Mandate - legal acts and other agreements
Data Sharing matters are governed by the Malta Statistics Authority Act XXIV of 2000.
All labour variables are sent to Eurostat.
7.1. Confidentiality - policy
Apart from the established legislative framework, the NSO is also guided by an internal general confidentiality policy. Data tagged as confidential are not available for publication.
7.2. Confidentiality - data treatment
Confidentiality matters are regulated by the Malta Statistics Authority Act XXIV of 2000. Such data are marked as such.
8.1. Release calendar
An advance release calendar is published on the NSOs website.
8.2. Release calendar access
All releases are published and disseminated at 1100 hrs as scheduled in the Advance Release Calendar. The calendar is published on the NSO website and includes a 3 month advance notice (the current month and the next two months).
8.3. Release policy - user access
National statistical releases are issued according to the release calendar and are made available to all users simultaneously. All releases are posted on the organisation's website.
Industrial labour variables are submitted to Eurostat 60 days after the end of the reference period on a monthly basis.
Construction labour variables are submitted to Eurostat 75 days after the end of the reference period on a quarterly basis.
Retail trade labour variables are submitted to Eurostat 75 days after the end of the reference period on a quarterly basis.
Services labour variables are submitted to Eurostat 60 days after the end of the reference period on a quarterly basis.
National news releases for these indicators are carried out on a quarterly basis and according to the national news release calendar.
10.1. Dissemination format - News release
The main dissemination channel for official statistics is the NSO website.
A quarterly news release is published on the industrial indicators and another news release covers retail and services indicators.
10.2. Dissemination format - Publications
The short-term statistics only publishes its figures in an annual compendium named Malta in Figures or Trends in Malta. These can be accessed from the NSO website.
10.3. Dissemination format - online database
The NSO website offers an online statistical database (StatDB) that provides a more in-depth analysis for the more expert statistical user or researcher. The StatDB requires registration as only registered users can extract data from the database. Registration is free-of-charge.
10.4. Dissemination format - microdata access
In general, index results made available to Eurostat ensure a sufficient degree of representativeness of data. More detailed data than that published locally and internationally are available on request but are subject to the confidentiality clauses.
In such cases, data are provided in index form. Micro data are not provided to users.
The NSO quality assurance principles are developed according to the ESS Handbook for Quality Reports (2014).
11.2. Quality management - assessment
Aspects of quality are divided into several steps. These are specifying needs, designing and building data instruments and collecting, processing, analysing, disseminating, archiving and evaluating data. The main guiding principles in quality management are found in the:
Overall, the quality of the labour indicators for all domains is reliable.
12.1. Relevance - User Needs
The main users of these indices are Eurostat, the Central Bank of Malta, the Ministry of Finance, Malta and other international organisations.
12.2. Relevance - User Satisfaction
Not available
12.3. Completeness
Data are compliant with relevant European and local regulations and requirements are fully fulfilled.
13.1. Accuracy - overall
The overall accuracy of the variables is considered to be of good quality.
Sampling error is considered to be small, non-sampling error is small and the first estimates are not biased.
13.2. Sampling error
Sampling error is considered to be small or not applicable because for many of the NACE categories, the sample designs are based on cut-off thresholds.
13.3. Non-sampling error
Non-sampling errors relate to problems such as under or over coverage, non-response, data-processing and estimation.
Non-sampling errors in the labour indicators are mainly due to non-response error. Hence, the Short-term Statistics unit make use of administrative data and good estimation techniques where needed. The hours worked variable may suffer a bit more from this problem than the other labour indicators. This is due to the relatively more difficult computational nature of the variable from the part of the units in the sample.
Administrative Registers help us to fill in a lot of non-response. Weighted response rate is over 75% using employment as the weighting variable.
14.1. Timeliness
Industry and services labour variables are released 60 days after the end of the reference period while construction and retail labour variables are released 75 days after the end of the reference period. Questionnaires are sent a week before the end of the reference period.
The data series has been compiled since the year 2000. The first set of indicators was based on NACE Rev 1. Following 2009, the indicator was published according to the new NACE Rev 2 classification. Furthermore, for every year ending in 0 or 5, new fixed aggregation weights are introduced.
There is no break in series in the data.
15.3. Coherence - cross domain
Employment data are sometimes compared to administrative data, however no other checks are made.
15.4. Coherence - internal
Aggregates may not always be consistent with sub-aggregates due to the nature of the index calculation.
Data burden is considered high but no data are available.
17.1. Data revision - policy
Revisions are carried according to the European Business Statistics Regulation. Internal procedures are also established.
The same revision policy is applied to both data released nationally and that sent to Eurostat.
Revision analysis is carried out before every news release published locally or sent to Eurostat and clear explanations are also sought. Detailed explanations at micro data level are also analysed but only used internally for quality information purposes. Generally, this index will have routine minor revisions due to actual data replacing estimates due to non-response. Major revisions may also be done but this will be on an exceptional basis.
A revision calendar does not exist.
17.2. Data revision - practice
Data marked as provisional are revised when actual data are received and estimates are replaced.
18.1. Source data
Data on the industrial labour variables are based on a sample survey of around 200 units.
Data on the construction labour variables are based on a sample survey of around 135 units.
Data on the retail trade labour variables are based on a sample survey of around 700 units.
Data on the services labour variables are based on a sample survey of around 800 units.
The samples are refreshed every 5 years and turnover and employment are the main indicators used to stratify the samples into 3 size strata. In certain NACE divisions, a cut-off of 80% was used for the Base 2015 data.
18.2. Frequency of data collection
Industry labour variables - monthly.
Construction, retail trade and services - quarterly.
18.3. Data collection
Data are collected by means of a monthly statistical questionnaire. Questionnaires are sent by post and their collection is done via post, fax, e-mail or online systems.
18.4. Data validation
Validations of format and file structure checks and intra-dataset checks are checked with the use of in-house excel sheets and access databases.
Revisions analysis is also carried out for every data transmission sent to Eurostat using in-house access queries. When data revisions are significantly large, a short explanation is also given to Eurostat.
18.5. Data compilation
Non-response is treated by micro-estimation to ensure a full response for all surveyed units.
A Laspeyres type index is used for compilation and the weights are updated every 5 years from the Structural Business Statistics survey.
The current file conversion is GESMES/TS and the plan is to use the SDMX-ML tool in the future.
18.6. Adjustment
Necessary working day and seasonal adjustments are made to the variables of hours worked and gross wages and salaries. No adjustments are made to the employment variable.
The software used for adjustments is JDemetra+ 2.2.0. The calendar used is country-specific and replacement of outliers is not done.
The direct method of adjustment is used for all the variables. Models are revised once every year and the whole time series is revised with every additional data.
To check the quality of the indicators, the Charts and S-I Ratio in the Main Results tab are always checked. Calendar effects are checked by analysing t tests, p values and outlier effect.
Seasonality is checked mainly by analysing the stability and seasonality tests (Friedman test, Kruskall-Wallis test, evolutive seasonality test, combined seasonality test and residual seasonality test) of the series. There are no seasonal breaks and filter length is chosen automatically.
Adjusted data is then compared with the previous transmission / news release to check for any large discrepancies.
This metadata file brings together all the labour variables (employment, hours worked and wages) of all the STS domains (industry, construction, retail trade and services).
The short-term statistics labour variables are the number of employees and self employed persons, hours worked by employees and gross wages and salaries.
These labour variables are compiled for the industry, construction, wholesale and retail trade and services domains.
However, in practice, the self employed persons sub-variable is not observed.
13 June 2025
The definitions of the main variables for the domains of industry, construction, trade and services are the following:
Employment: The number of employees includes all paid and unpaid workers. Self-employed persons are also included.
Hours worked by employees: Hours worked includes the total amount of hours actually worked, that is: regular working hours and also during overtime; during nights, Sundays or public holidays. Hours worked but not paid are also considered. Hours worked does not include vacation and sick leave, loss of work-hours due to accidents, strikes, lock outs, slack time etc. and also time spent for meal breaks and commuting strike days. The hours worked for self-employed persons are not included.
Gross wages and salaries:Gross wages and salaries are defined as the total sum of remuneration in cash or in kind payable to all persons employed in return of work done during the reference period, irrespective of whether this remuneration is paid regularly or not. Direct taxes and employees’ social security and pension contributions are included. Wages and salaries of self-employed persons are not included.
The statistical unit is the Kind of Activity Unit (KAU).
Industry variables: The statistical population is around 2000 units and includes all those units falling under NACE divisions 8 - 36.
Construction variables: The statistical population is around 3300 units and includes all those units falling under NACE divisions 41 - 43.
Wholesale and retail trade variables: The statistical population is around 9000 units and includes all those units falling under NACE divisions 45 - 47.
Services indicators variables: The statistical population is around 6000 units and includes all those units falling under NACE sections H, I, J, L, M and N.
The indicators cover the national territory of the Maltese Islands. They do not cover activities outside this geographical area.
Industrial labour variables are submitted on a monthly basis.
Construction labour variables are submitted on a quarterly basis.
Retail trade labour variables are submitted on a quarterly basis.
Services labour variables are submitted on a quarterly basis.
The overall accuracy of the variables is considered to be of good quality.
Sampling error is considered to be small, non-sampling error is small and the first estimates are not biased.
The data are provided in index form both to Eurostat and when publishing news releases on a national level. Percentages changes from one period to another are also published.
Non-response is treated by micro-estimation to ensure a full response for all surveyed units.
A Laspeyres type index is used for compilation and the weights are updated every 5 years from the Structural Business Statistics survey.
The current file conversion is GESMES/TS and the plan is to use the SDMX-ML tool in the future.
Data on the industrial labour variables are based on a sample survey of around 200 units.
Data on the construction labour variables are based on a sample survey of around 135 units.
Data on the retail trade labour variables are based on a sample survey of around 700 units.
Data on the services labour variables are based on a sample survey of around 800 units.
The samples are refreshed every 5 years and turnover and employment are the main indicators used to stratify the samples into 3 size strata. In certain NACE divisions, a cut-off of 80% was used for the Base 2015 data.
Industrial labour variables are submitted to Eurostat 60 days after the end of the reference period on a monthly basis.
Construction labour variables are submitted to Eurostat 75 days after the end of the reference period on a quarterly basis.
Retail trade labour variables are submitted to Eurostat 75 days after the end of the reference period on a quarterly basis.
Services labour variables are submitted to Eurostat 60 days after the end of the reference period on a quarterly basis.
National news releases for these indicators are carried out on a quarterly basis and according to the national news release calendar.
Industry and services labour variables are released 60 days after the end of the reference period while construction and retail labour variables are released 75 days after the end of the reference period. Questionnaires are sent a week before the end of the reference period.
The data series has been compiled since the year 2000. The first set of indicators was based on NACE Rev 1. Following 2009, the indicator was published according to the new NACE Rev 2 classification. Furthermore, for every year ending in 0 or 5, new fixed aggregation weights are introduced.
