National Statistics Office (NSO)

Unit C2: Labour Market Statistics

Directorate   C - Social Statistics and Information Society

Malta Statistics Authority
c/o National Statistics Office
Lascaris
Valletta VLT 2000
Malta

The LCS 2012 was based on Commission Regulation (EC) No 1737/2005 of the 21^st  October 2005 amending Regulation (EC) No 1726/1999 as regards the definition and transmission of information on labour costs.  In this regard all compulsory variables were covered by the survey and the target population was enterprises engaged in NACE Rev 2 Sections B to N and P to S and employing 10 or more employees.

This report is intended to cover the following items as per Commission Regulation 698/2006 of 5^th May 2006 implementing Council Regulation (EC) No 530/1999 as regards quality evaluation of structural statistics on labour costs and earnings:

1. Relevance
2. Accuracy
1. Sampling errors
2. Non-sampling errors
3. Punctuality and Timeliness
4. Accessibility and clarity
5. Comparability
6. Coherence

Not available.
New concept added with the migration to SIMS 2.0.
Information (content) will be available after the next collection.

Not available.
New concept added with the migration to SIMS 2.0.
Information (content) will be available after the next collection.

Not available.
New concept added with the migration to SIMS 2.0.
Information (content) will be available after the next collection.

Not available.
New concept added with the migration to SIMS 2.0.
Information (content) will be available after the next collection.

Not available.
New concept added with the migration to SIMS 2.0.
Information (content) will be available after the next collection.

Not available.
New concept added with the migration to SIMS 2.0.
Information (content) will be available after the next collection.

Not available.
New concept added with the migration to SIMS 2.0.
Information (content) will be available after the next collection.

Not available.
New concept added with the migration to SIMS 2.0.
Information (content) will be available after the next collection.

-

Sampling methodology

The Sampling Frame

The sampling frame for this survey was the Business Register which is maintained by the Business Statistics Unit within NSO.  This register contains data regarding legal units which are recognized as having autonomous management and an independent accounts system at NUTS 1 level.  In this regard the target population for LCS could be chosen from this database.

The total number of enterprises operating in NACE Rev. 2 Sections B to N and P to S and employing 10 or more employees amounted to 2110. 

The statistical unit for this survey was the enterprise.  Even though the definition of a local unit is different from that of an enterprise, one can safely say that in a local context this matter should not be an issue given the small size of our country.  In addition, the sampling frame which was used for the extraction of the LCS sample is based on enterprises rather than local units and hence the LCS statistical unit had to be the enterprise as well.

According to Commission Regulation (EC) No 1737/2005, LCS data has to be collected for enterprises operating in NACE Rev. 2 Sections B to N and P to S split in the following size classes:  10 to 49, 50 to 249, 250 to 499, 500 to 999 and 1000+. 

Sample selection was made using stratification by optimal allocation.  The NACE sections and employment groups (as per Appendix 2 in Regulation No. 1737/2005) were used in order to create the strata.  

For national purposes and for a better and more adequate representation, the first size class (10 to 49) was split in 10 to 19 and 20 to 49.  

After stratifying enterprises by NACE and size class a sample of 1255 companies was chosen.  Table A, Table B and Table C (in annex) illustrate the distribution of the population, net sample distribution and the probability of selection for each strata.

Every four years

The Office instructed the IT unit to devise an application which could enable the automatic emailing of questionnaires to the different respondents along with pre-filled details of the enterprise. The questionnaire was made available in excel version since this software is widely available and since it facilitates the copying and dragging of information for different aspects of this survey.  In cases when emails of enterprises were not available, the questionnaires were sent by post.

To minimize processing errors, each incoming questionnaire is thoroughly checked by trained statisticians using a number of validations. In addition more checks are carried out at the analysis stage whereby data is compared on a per capita basis in order to determine any incorrect or inconsistent information.

Prior to the compilation of weights, checking relating to economic activity and size class of enterprise is carried out in order to ensure that sampled units are in line with the target population.  Then, the post stratification weights are  based on following variables: economic activity at section level, and size of enterprise.

Below is a description of how imputations for different variables were carried out.

A1 Number of employees

Mathematical imputation of this variable was not required since all item non-response concerning number of employees was collected using auxiliary information from other enterprise-based surveys held by NSO.  For a number of cases where such information was not available from other surveys, the responding unit was contacted directly to provide this missing data. 

B1 Hours actually worked and C1 Total hours paid

The national questionnaire asked respondents to estimate the sum of hours actually worked (including overtime hours) as well as estimates for the following variables for the year 2012:

• Hours of vacation leave and public holidays
• Sick leave and injury leave hours
• Hours for maternity leave
• Other paid hours

In addition, respondents were asked to provide the average number of hours worked per week for full time, part time and apprentice workers.

Figures on actual hours worked and on total hours paid were divided by 52 (to reflect the hours worked per week) and were split on a per capita basis.  The answer was cross checked with information on the normal hours worked per employee per week.  Whenever the sum of hours actually worked was equal to the normal hours worked, figures were tagged for subsequent imputation since the assumption was that respondents were not taking into account the number of hours which were paid and not worked in the actual hours worked computation.

Hot-deck imputation using median actual hours worked per employee by economic activity or size class was used in order to cater for item non-response related to this variable. 

The paid hours not worked were imputed using the median figures for the economic activity or size class of the sampling unit.  Imputations for actual hours worked for full time employees and apprentices were conducted by deducting the paid hours not worked from the estimated total normal hours worked. In the case of actual hours worked for part timers, the median per capita value per economic activity or size class was used for imputing missing or inconsistent data.

D1  Labour costs

The national questionnaire contained a breakdown of labour costs.  Respondents were asked to provide information on expenditure related to the following variables:

1.1   Wages and salaries

1.2   Overtime payments

1.3   A breakdown of bonuses not paid in each pay period

i. Statutory bonuses
ii. Production bonuses paid irregularly
iii. Performance bonuses
iv. Leaving and retirement bonuses
v. Other bonuses

1.4   A breakdown of bonuses paid in each pay period

i. Shift allowances
ii. Sales commissions
iii. Danger money
iv. Production bonuses paid regularly

1.5   Payments in kind

1.6   Statutory social security contributions

1.7   Other non-obligatory social contributions

1.8     Sick leave payments incurred by the employer

1.9    Maternity leave payments incurred by the employer

1.10     Injury leave payments incurred by the employer

1.11     Payments to employees leaving the enterprise

1.12     Payments to health insurance schemes

1.13     Payments to early retirement schemes

1.14     Payments for study grants

1.15     Other social contributions

1.16     Vocational training costs

1.17     Other expenditure

1.18     Subsidies received by the employer in connection with labour costs

Imputations were therefore carried out at this level.

For the 2012 version the LCS questionnaire included a list of questions in order to determine whether a specific cost was applicable for the sampling unit or not so that imputations are only applied wherever applicable.

1.1   Wages and salaries

There were very few cases which needed to be imputed in this case. Such cases were imputed using the overall median hourly rate by economic activity and size class. No other imputations were necessary for this variable since corrections made were the result of processing errors.

1.2   Overtime payments

Figures for imputation were identified by comparing the annual basic wage per capita with the annual overtime pay per capita and by comparing the basic hourly rate with the overtime rate.  In addition cross checks between overtime hours worked and overtime payments were made in order to identify cases where overtime hours were given but no corresponding figure was provided for overtime payments.  Imputations were subsequently based on the median overtime hourly rate by economic activity.

1.3   Bonuses not paid in each pay period

Data for this variable was compared to data which was collected in 2008.  Any large variations were identified for further checking.  Bonuses per capita values were also analyzed. Where imputations were necessary, they were based on the median bonuses not paid in each pay period per employee by size class or by NACE group in FTE.

For cases in which information for 2012 LCS was not available, figures where checked for their consistency, usually using per capita rates and imputations were based on median bonuses not paid in each pay period per employee by size class or by NACE group in FTE.

Statutory bonuses

Missing statutory bonuses were imputed based on the number of full time and part time workers since the amount of statutory bonuses incurred by an employer differs depending on the type of employment.

Performance bonuses

Data was checked at micro level and compared to information which companies forwarded in the LCS 2008.  Where imputation was necessary, this was carried out using per capita values by NACE or size class of the company.

1.4   Bonuses paid in each pay period

As above, amongst other checks and validations, data for this variable was compared to information which was collected in 2008.  Any large variations were identified for further checking. Bonuses per capita values were also analyzed. Where imputations were necessary, they were based on the median bonuses paid in each pay period per employee by size class or by NACE group.

Shift allowances

Data for this variable was checked by referring to the 2008 LCS. In such cases, NSO directly phoned respondents to check whether such payments were still applicable for 2012 and subsequently imputed information using 2008 figures and adding the percentage increase in salaries (cost of living increase) between 2008 and 2012  data.

Sales commissions

In order to identify whether sales commissions were a labour cost for specific economic categories, the mean per employee for each variable was worked out from data provided in the LCS.  In this way, one could identify which economic activities were likely to have such payments.  Subsequently data at micro level was referred to in order to identify which NACE activities needed imputations.  Data for this variable was also checked by referring to the 2008 LCS.

Imputations were based using the FTE employment for the companies with missing data multiplied by the median per capita by economic activity.

Production bonuses

Data was analysed at micro level and companies which were likely to have production bonuses were either contacted directly to supply this missing information or data was imputed using the per capita value by NACE or size class

1.5   Payments in kind

Per capita payments in kind were worked out on existing information and extremes in data were checked with the data provided in the questionnaire by respondents.  Comparisons with 2008 LCS data were also carried out in order to identify companies which failed to provide such data in 2008 and imputations were carried out based on the median per capita value per economic activity and size class.

1.6 and 1.7  Statutory and non-obligatory social security payments

Checks for this variable concentrated on comparing the percentage of statutory social security payments out of total wages and salaries.  This percentage was compared with 2008 figures so that any large variations were checked with respondents or with other data which was available in the questionnaire.  Large changes between 2008 and 20012 were first looked into by checking variations in employment in both years, since, such changes would result in differences in statutory social security contribution payments.  Imputations were made by referring to the mean full time equivalent employment and the per capita value of this cost as a percentage of the total wages and salaries within the economic activity and size class of the unit.

1.8, 1.9 and 1.10   Sick, maternity and injury leave payments

Respondents were asked to provide information on the number of sick, maternity and injury hours, availed of by employees during the reference year 2012.  This information helped in estimating missing information on the amounts of sick, maternity and injury payments.  Imputations were based on the basic hourly rate (worked out by dividing the basic wages and salaries by the sum of hours excluding overtime).  This rate was in turn used to impute maternity and injury payments by the total number of sick, maternity and injury leave hours.

1.11   Payments to employees leaving the enterprise

Checks were carried out for companies which recorded a decrease in employment between 2008 and 2012 and also during 2012 itself.  Companies which were identified were checked at a micro level using information available from employment data from administrative records.  Where necessary, confirmations with respondents were also carried out in order to check if information on payments to employees leaving the enterprise were applicable to the sampled unit. 

1.12   Payments to health insurance schemes

Imputations were based on the median per capita value by economic activity multiplied by the number of full time employees.

1.13   Payments to early retirement schemes

Payments for this variable were divided by the number of full time and part time employees in order to have a per capita value. Exceptionally high values where identified for additional checking with respondents.  In cases where it was common knowledge that early retirement payments had taken place because of the restructuring of public sector companies, information was retrieved from information available within the Office’s government finance unit.

1.16  Vocational training costs

Imputations for vocational training costs were based on the median per capita value multiplied by the full-time equivalent employment by economic activity.

1.17  Other expenditure

Imputations for other expenditure were based on the median per capita value multiplied by the full-time equivalent employment by economic activity.

D4 Taxes

No imputations were necessary for this variable since employers do not have any taxes related to labour costs.

1.18  Subsidies received by the employer

This check was mainly focused on NACE Rev 2 Sec P since at a national level schools run by the Church have their labour costs subsidized by Government.  Other imputations were necessary for companies which were employing apprentices under specific schemes run by the national employment agency.  In this regard use of administrative records helped to identify companies which were benefiting from any apprentice-related schemes.  The Planning and Priorities Coordination Department also provided information on the payments which these companies were receiving as part of the scheme.  This information was subsequently used to impute missing data.

Not available.
New concept added with the migration to SIMS 2.0.
Information (content) will be available after the next collection.

[Not requested]

-

No user survey to determine the needs of LCS users has been carried out.   From information available regarding the actual use of this survey, one can say that LCS data is mainly used for the computation of the indicator on annual labour costs and to validate results obtained from other enterprise based surveys carried out by the NSO.

One can point out that since this survey is carried out once every four years, users might not find it to be very useful, given that they usually require information on a more regular basis.

All compulsory variables requested in Commission Regulation 1737/2005 were covered in the 2012 Labour Cost Survey.

[Not requested]

-

Great effort was made in order to ensure that highest quality data was collected in this survey.  In this regard, employers and human resource managers were encouraged to provide information using internal data from databases or registers.  However, whenever such data was not available because it was not compiled, employers were encouraged to provide estimates.  NSO believes that since employers have a better and more in depth knowledge of their company, any estimates are bound to be of a better quality if provided by respondents themselves. 

Having said this, if data provided was not deemed to be of sufficient quality or did not make sense when compared to other information available (e.g. data from business register), employers were contacted to clarify estimates and figures provided.  If this was not possible, then misleading data was deleted and imputed.  

Kindly refer to 5.2.1

This section of the report contains a number of tables in annex on the coefficient of variation for the following variables:

• NACE
• Size-class

Probability sampling has been used for LCS and therefore there are no non-probability sampling errors.

1. Misclassification errors

Coverage errors which have been identified in LCS include errors related to misclassification and other coverage errors.

 1.1 Misclassification of NACE

Misclassification errors refer to incorrect NACE classification to units present in the target population.  In this regard, no enterprises have been misclassified.

1.2 Misclassification of size class 

Another aspect of misclassification concerns size class.  Table 3 indicates the extent of this misclassification.  A comparative analysis is provided showing the misclassified units by size class prior to data collection and after information was returned to the office. In total, 54 companies have been misclassified. The reasons for such a misclassification were mainly two:

• enterprise employed less than 10 employees; or
• enterprise was no longer existent following a merger or acquisition in the reference year  2012.

As a result these companies had to be removed from the LCS 2012 sample since they were not part of the target population.

Table: Misclassification of companies by Size Class

* misclassified either because enterprise employed less than 10 employees; or it was liquidated/closed down; or merged with another enterprise in the reference year 2012.

Other units which had been assigned a different size class prior to the data collection were reclassified after the data collection phase was concluded and weights were worked out accordingly.

Over-coverage and under-coverage errors

Over-coverage errors found in LCS mainly related to misclassified units which were not within the scope of the survey or units which were no longer active during the reference period identified for the LCS.  A total of 54 units were excluded from the survey because of these reasons.

The main reason which can be attributed to the over-coverage errors was the time lag between the reference period and the actual data collection.  Business register data for 2012 provided information on the number of active units at the time and the economic activities and employment levels of the units.  Data collection however took place 6 months after the end of the reference period and during this time units might have changed their activities, reduced the number of employees or ceased to operate altogether, thus making themselves ineligible for the survey.

Under-coverage problems were tied to the following reasons:

• enterprises for which the sampling frame registered an employment of 9 or less employees, whereas in actual fact they had 10 or more persons on their payroll, and
• enterprises which were established in 2012 but which were not yet included in the sampling frame

In order to account for problems of under and over coverage, the Labour Market Statistics unit had to ensure that the sampling frame was updated before the actual weighting procedure was carried out.  This correction however resulted in changes in the original probability of selection for some strata.

[Not requested]

Bias is attributed to all forms of human error committed during the data collection and analysis stage, despite all the efforts that were made by the NSO in order to ensure that highest quality data is collected and compiled for this survey.  These may include errors committed by the respondents during the filling of the questionnaire (e.g. under-reporting, respondents not following properly the instructions set in the questionnaire, etc.). 

Unfortunately, this survey could not be conducted using face-to-face interviewing,  due to lack of resources that were available. Given the ever increasing need for making data collection a more simple process, the Office instructed the IT unit to devise an application which could enable the automatic emailing of questionnaires to the different respondents along with pre-filled details of the enterprise. The questionnaire was made available in excel version since this software is widely available and since it facilitates the copying and dragging of information for different aspects of this survey.  In cases when emails of enterprises were not available, the questionnaires were sent by post.

Other errors include data entry errors as well as errors committed during the vetting and analysis stages (e.g. imputation).  The effect of non-response cannot be ignored since significant non-response in particular strata might lead to misleading results.

NSO tried to minimize measurement errors during different stages of the data collection.

1. Accounting for measurement errors from the questionnaire

NSO’s initial but very important objective was to have a questionnaire which was   self-explanatory to respondents without creating excessive response burden.  In fact, NSO took on board feedback about the structure and content of the LCS questionnaire from past LCS waves. The questionnaire was modified where necessary, in order to have a clearer and user friendly version.

Respondents were also provided with additional assistance by staff working within the Labour Market Statistics Unit. Such assistance was mostly provided via telephone and email. 

2. Accounting for Respondent errors

Since LCS data had to be filled in by respondents themselves, NSO tried to make questions comprehensible so that even if the respondent refrained from reading supporting explanatory notes, s/he would still answer correctly. 

In terms of errors resulting from lack of information on respondents’ registers or databases, respondents were asked to provide estimates.  In this regard, feedback from respondents shows that the following variables were found to be difficult to retrieve from companies’ records:

• Actual hours worked
• Paid hours not worked
• Sick/Maternity/Injury Leave Payments incurred by the employer

NSO tried to minimise respondent burden by trying to collect data in a way which was easier for the respondent to provide and understand.  Terminologies which are being used at a national level where included in the questionnaire to facilitate the filling in of the survey.

NSO asked additional questions in order to be in a position to check on the provided information.  As done in the previous wave of the LCS,  respondents were asked to provide the average number of weekly hours normally worked for full, part time employees and apprentices since this information was very useful for checking the answers provided on actual hours worked. Moreover, when LCS 2008 data was available for the same enterprise, information was crossed checked at a micro level.   Internal records available within the Short Term Business Statistics unit were also referred to for checking the consistency of the information which was being provided. In the LCS 2012 version, respondents were asked beforehand to state whether the enterprise incurred any labour costs during the year. The main aim of these questions was to validate the answers with other questions relating to labour costs asked throughout the survey.   

Refer to 5.3.3.1 and 5.3.3.2.

The total number of units contacted for the survey was 1281, of which 1187 resulted to be eligible for the survey. A total of 631 of the eligible units completed the LCS questionnaire.  The response was subsequently improved using Short-term Business Statistics data along with information kept at the Malta Financial Services Authority for an additional 203 units.  This means that the response rate for responding units was  53.2% ( 631 out of  1187 units) whilst the inclusion of units using other sources, added a further 17.1% to the overall response rate (203 out of  1187 units).  Hence the total unit response rate for the Labour Cost Survey 2012 stood at  70.3% (834 out of 1187  units).

Information on the item imputation rates for all mandatory variables is illustrated in the table in annex "Item imputation rates".  

To minimize processing errors, each incoming questionnaire was checked by trained statisticians using a number of validations. 

Any data inconsistencies were then verified directly with the sampling unit’s contact person either by phone or by email.  To minimise processing errors, the office asked for the details of two contact persons to ensure accountability for each and every figure reported in the survey.

The overall imputation rate resulted to be 14.9%. This rate is based on all the mandatory variables.

I.   Ensuring the good coverage of the target population

In order to ensure that the target population of companies was well covered, stratified random sampling was applied for those companies operating in NACE Rev  2 sections B to N and P to S employing 10 or more persons.  Moreover, due to national requirements, the 10 to 49 size class was split into two categories, namely, 10 to 19 and 20 to 49.  This was done in order to ensure a better and more adequate representation of data for the local scenario.

II.   Combination of data from administrative and other sources

NSO made use of administrative data compiled by tax department, enterprise annual reports housed at the national financial services authority and short term business statistics data relating to the year 2012. 

[Not requested]

[Not requested]

[Not requested]

[Not requested]

[Not requested]

-

LCS data was transmitted to Eurostat on the 30^th June 2014 and following error checks by Eurostat a final dataset was provided on the 31^st July 2014.  

[Not requested]

[Not requested]

The table in annex ("Data collection and post collection phases") illustrates the various stages between data collection and analysis.

From this table, one can note that considerable time was taken to complete data collection. This was mainly due to the fact that key companies which had not forwarded their data and which had a probability of selection of one had to be chased numerous times until they submitted their questionnaire.  The small size of the country necessitates the use of a census for most business surveys which are carried out amongst companies employing 50 or more employees.  This fact therefore places a huge response burden on selected organizations which are constantly being contacted for various surveys which are held by various units within NSO and also by other entities in general.   Since a number of these organizations are key players in their respective sectors one cannot afford to leave them out of any business survey since most often they determine the developments which are taking place within the sector in which they operate and consequently influence the representativity of the results

The above issues also had a direct effect on the timings of data analysis and weighting procedures.

[Not requested]

-

-

National concepts applied for LCS are in line with European concepts since the definitions outlined in Commission Regulation 1737/2005 were applied in the local context. 

In terms of the statistical units which were covered for LCS, data was collected from legal units which are recognized as having autonomous management and an independent accounts system.  At NUTS 1 level the whole country is represented therefore information could be collected from enterprises which were recognized to be legal units by the Business Register.

[Not requested]

LCS 2012 results were compared with LCS 2008 both at macro-level and even at a micro-level when available. Between both periods there was no deviation from the regulation, hence data can be compared.

[Not requested]

1. Coherence with the Labour Force Survey

In order to ensure adequate comparability of LCS data with LFS data, information from the latter survey had to be customized.  In this regard, Labour Force Survey data for four different quarters in 2012 was combined into one dataset and a selection of employees with a main job in companies with 10 or more employees and operating in NACE Rev. 2 sections B to N and P to S was made.

In annex one can find the actual hours worked data in the reference year which have been obtained from the LFS compared with those obtained from the LCS.

When analyzing data on employment by size of enterprise from the LFS, one is to be careful about the interpretation of results, since the LFS is not designed to come up with a proper estimate of the number of units employing 10 or more employees.   In addition, LFS provides the household perspective, whereas LCS gives the enterprise perspective.  Hence there is bound to be differences in the results obtained. 

Another factor which is bound to influence the results relates to the fact that LFS has more rigorous checks on hours worked since information refers to a specific week and this can be better checked at the data validation stage.  By contrast, information from LCS on hours worked refers to the annual amounts and hence it is likely that figures are less accurate.

The information presented for LFS refers to the main job as per EC Regulation 698/2006, whereas LCS data includes all hours worked by employees irrespective of whether their employment is a main or a second job.   This factor is bound to produce higher averages for LFS data, since persons working a second job are more likely to work less hours and hence their inclusion would drive the average hours downwards.  

2. Coherence with Structure of Business Survey data

Information from the Structure of Business Survey was retrieved using the variable entitled ‘Gross wages and salaries’ whereas information from LCS was worked out using a range of variables. 

The variations between the two surveys are the result of the restricted details which emanate from SBS.  The use of one variable in SBS to estimate annual wages and salaries is likely to give different results when compared to LCS data which is highly more focused on this element of labour costs. NACEs K,P,Q,R and S are not surveyed by the SBS Unit and hence no comparisons at this level can take place. 

[Not requested]

The table in annex illustrates the growth rates from LCI (based on NACE Rev 2) and LCS (based on NACE Rev 1.1) for NACE aggregations. Although having 2 different NACE classifications, this comparison is still possible since, at this aggregate level, they are very similar to each other as regards economic activities.

Variations between LCI and LCS figures, are mainly the result of the micro business effect (under 10 effect) which is being taken into account in the LCI averages but is missing in the LCS averages. This is the result of lower growth rates in case of the LCI when compared to those in the LCS.

National Accounts data is being compared to LCS data.  One is to note however that National Accounts information relates to all companies operating in the sector whereas LCS data refers to companies which employ 10 or more employees.

Variations between National Accounts and Labour Cost Survey figures, are the result of the micro business effect (under 10 effect) which is taken into account in the National Accounts averages but is missing in the LCS estimate. 

[Not requested]

At a national level, results are intended to be published during 2015.  These results will be published in the form of a news release or as part of a publication encompassing labour-market-related statistics and will be disseminated to the media and via the office’s website.  Methodological notes will be accompanying the news release data.

A news release in March 2015 has been published on a national level.

LCS statistics are not included in national publications.

Can be accessed through: http://nso.gov.mt/statdb/start

[Not requested]

[Not requested]

Aggregated tabular data on LCS statistics can be provided to users who request this type of information. Such requests can be made on the NSO’s website through: http://nso.gov.mt/en/Services/Pages/Request-for-Information.

Documentation of steps on LCS is available for internal purposes. In addition, this Office uses the implementation arrangements provided by Eurostat: EC 530/1999 and EC 1737/2005.

[Not requested]

[Not requested]

[Not requested]

[Not requested]

-

At National level: Confidentiality is one of the major principles guiding the activities of the NSO.

Article 40 of the MSA Act stipulates the restrictions on the use of information and in Article 41, the prohibition of disclosure of information. Furthermore, Section IX of the Act (Offences and Penalties) lays down the measures to be taken in case of unlawful exercise of any officer of statistics regarding confidentiality of data. No cases of breaches in the law have been recorded to date.

Since its inception, the NSO has always operated within a culture of strict confidentiality to which it is also bound by the provisions of the Data Protection Act of 2000. This Act, which came fully into effect on July 15, 2003, seeks to protect individuals against the violation of their privacy by the processing of personal data.

Refer also to the NSO policy "Confidentiality of Personal and Commercial Data" - http://nso.gov.mt/en/nso/About_NSO/Documents/NSO_Policies/Confidentiality_of_personal_and_commercial_data.pdf

Also further information on access to microdata is available on the NSO's website through: http://nso.gov.mt/en/Services/Microdata/Pages/Access-to-Microdata.aspx

During 2009, the NSO has set up a Statistical Disclosure Committee to ensure that statistical confidentiality is observed, especially when requests for microdata are received by the NSO.

Upon employment, staff is informed of the rules and duties pertaining to confidential information and its treatment. According to the MSA Act, before commencing work, every employee is required to take an oath of secrecy whose text is included in the Act.

At European level: Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4) of 11 March 2009 (OJ L 87, p. 164), stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.

For the LCS 2012, primary and secondary confidentiality are taken into consideration.  Primary confidentiality is flagged on too few enterprises, where the number of units is less than 3 and the dominance share of the largest units is 75%. Secondary confidentiality data is flagged in order to protect primary confidential data, which are suppressed so that sensitive information is not revealed. They are identified and flagged by NSO using a common methodology applied by other statistical agencies.

No further comments.