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.
Foreign Affiliates Statistics (FATS) measure the commercial presence through affiliates in foreign markets.
Inward foreign affiliates statistics (IFATS) shall mean statistics describing the activity of foreign affiliates resident in the compiling country.
In country-level business statistics foreign-controlled enterprise shall mean an enterprise resident in the compiling country over which an ultimate controlling institutional unit not resident in the compiling country has control.(Table 14 of the EBS Implementing Regulation (EU) 2020/1197).
Variables on the country-level business activities in the IFATS data category:
Business activities in foreign control:
210301. Number of foreign-controlled enterprises
220501. Number of employees and self-employed persons in foreign-controlled enterprises
220701. Employee benefits expense in foreign-controlled enterprises
230301. Intramural R & D expenditure in foreign-controlled enterprises
230401. R & D personnel in foreign-controlled enterprises
240301. Total purchases of goods and services of foreign-controlled enterprises
240302. Purchases of goods and services for resale of foreign-controlled enterprises
250601. Net turnover of foreign-controlled enterprises
250701. Value of output of foreign-controlled enterprises
260201. Foreign-controlled enterprises’ gross investment in tangible non-current assets
250801. Value added of foreign-controlled enterprises
Business activities in total economy:
210101. Number of active enterprises
220101. Number of employees and self-employed persons
220301. Employee benefits expense
230101. Intramural R & D expenditure
230201. R & D personnel
240101. Total purchases of goods and services
240102. Purchases of goods and services for resale
250101. Net turnover
250301. Value of output
250401. Value added
260101. Gross investment in tangible non-current assets
3.2. Classification system
Classification systems used in the FATS are as follows:
Statistical classification of economic activities in the European Community (NACE Rev. 2);
List of 2-digit country codes (ISO 3166-1);
Currency codes (ISO 4217).
3.3. Coverage - sector
For all variables except for variables Intramural R & D expenditure, Intramural R & D expenditure in foreign-controlled enterprises, R & D personnel and R & D personnel in foreign-controlled enterprises: Market producers of NACE Sections B to N and P to R and divisions S95 and S96;
For variables Intramural R & D expenditure, Intramural R & D expenditure in foreign-controlled enterprises, R & D personnel and R & D personnel in foreign-controlled enterprises: Market producers of NACE Sections B to F.
3.4. Statistical concepts and definitions
Foreign Affiliates Statistics (FATS) measure the commercial presence through affiliates in foreign markets.
Inward foreign affiliates statistics (IFATS) shall mean statistics describing the activity of foreign affiliates resident in the compiling country.
In country-level business statistics foreign-controlled enterprise shall mean an enterprise resident in the compiling country over which an ultimate controlling institutional unit not resident in the compiling country has control.(Table 14 of the Implementing Regulation (EU) 2020/1197).
Foreign affiliate in the framework of outward FATS is an enterprise or branch not resident in the compiling country over which an institutional unit resident in the compiling country has ultimate (direct or indirect) control.
Domestic affiliate shall mean an enterprise resident in the compiling country over which a UCI resident in the same compiling country has control.
Ultimate Controlling Institutional of a foreign affiliate (UCI) shall mean the institutional unit, proceeding up a foreign affiliate’s chain of control, which is not controlled by another institutional unit.
Control is the ability to determine the general policy of the affiliate by choosing appropriate directors, if necessary. In this context, enterprise A is deemed to be controlled by an institutional unit B when B controls, whether directly or indirectly, more than half of the shareholders' voting power or more than half of the shares.
Indirect control means that an institutional unit may have control through another affiliate which has control over enterprise A.
Active enterprise is a statistical units which at any time during the reference period was ‘enterprise’, as defined in Regulation (EEC) No 696/93, and also active during the same reference period. A statistical unit is considered to have been active during the reference period if, in said period, it either realized positive net turnover or produced outputs or had employees or performed investments.
Employees and self-employed persons are persons who work for an observation unit on the basis of a contract of employment and receives compensation in the form of wages, salaries, fees, gratuities, piecework pay or remuneration in kind; and persons who are the sole owners or joint owners of the statistical unit in which they work. Family workers and outworkers, whose income is a function of the value of the outputs of the statistical unit, are also included.
Employee benefits expense contains all expenses arising in relation with employee benefits, recognized by the statistical unit during the reference period. Those are are all forms of consideration given by the statistical unit in exchange for service rendered by employees or for the termination of employment.
Research and experimental development (R & D) comprise creative and systematic work undertaken in order to increase the stock of knowledge – including knowledge of humankind, culture and society – and to devise new applications of available knowledge. Expenditures on intramural R & D represent the amount of money spent on R & D that is performed within a reporting unit. Intramural R & D expenditures are all current expenditures plus gross fixed capital expenditures for R & D performed within a statistical unit during a specific reference period whatever the source of funds. R & D current expenditures include labour costs for internal R & D personnel and other current costs (costs for external R & D personnel, purchase of services.). Gross fixed capital expenditures for R & D include: acquisition of land, acquisition of buildings, acquisition of information and communication equipment, acquisition of transport equipment, acquisition of other machinery and equipment, acquisition of capitalised computer software, acquisition of other intellectual property products.
R & D personnel in a statistical unit include all persons engaged directly in R & D, whether employed by the statistical unit or external contributors fully integrated into the statistical unit’s R & D activities, as well as those providing direct services for the R & D activities (such as R & D managers, administrators, technicians and clerical staff).
Total purchases of goods and services contains all amount of goods and services purchased by the statistical unit, recognized in accounting as either current assets or expenses during the reference period.
Purchases of goods and services for resale in are purchases of goods for resale to third parties without further processing. It also includes purchases of services by ‘invoicing’ service companies, i.e. those whose turnover is composed not only of agency fees charged on a service transaction (as in the case of estate agents) but also the actual amount involved in the service transaction, e.g. transport purchases by travel agents.
Net turnover consists of all income arising during the reference period in the course of ordinary activities of the statistical unit, and is presented net of all price reductions, discounts and rebates granted by it.
Value of output represents the value of the total output of the statistical unit, generated during the reference period.
Value added is a composite indicator of net operating income, adjusted for depreciation, amortization and employee benefits, all components being recognized as such by the statistical unit during the reference period.
Gross investment in tangible non-current assets includes all additions to tangible non-current assets, recognized as such by the statistical unit during the reference period, except any increases from revaluations or reversals of previously recognized impairment losses and from reclassifications (transfers) of other tangible non-current assets.
Description of the coverage of Special Purpose Entities (SPE)
SPEs are dealt with in three ways:
If the SPE is a "resident" unit under control, the SPE is dealt with during SBS data production. Any SPE present or absent in the SBS population is present or absent in the IFATS population;
Luxembourgish SPEs appointed as ultimate controlling institutional unit (UCI) are subject to validation, if they can be detected based on the thresholds or if they are classified under NACE code K 64.2 or 64.9. By definition, SPEs have little substance in the host economy and are therefore rarely appointed as controlling units. The definition of control suggests sufficient economic substance in the host economy. Control is also required to be active (i.e. local) rather than passive (i.e. remote-controlled).
Foreign SPEs appointed as UCI are accepted as such, unless we have sufficient data to prove that the UCI should be in a different country. UCIs assigned to offshore countries may hint at such SPEs.
Description of the treatment of equally shared control
We distinguish 5 types of equally shared control, for which we apply the geographical proximity principles:
control shared between one national (LU) and one foreign unit which are not linked by an enterprise group (i.e. both units being independent from each other): in this case we assign ultimate control to the the resident unit itself, the UCI country thus being under national control;
control shared between two independent units resident in the same country: the UCI country code is allocated to the said country of residence, no matter whether the country is LU (national), EU or extra-EU;
control shared between one country belonging to the European Union and one non-EU country, both units being independent from each other: the resident unit is under equally shared foreign control. Given the presence of a EU country, the UCI country code is allocated to the country of residence in the EU (dominant controller by assumption);
control shared between two different countries belonging to the European Union, both units being independent from each other: the resident unit is under joint foreign control, therefore the UCI country code is allocated to the relevant residual EU code;
control shared between two different countries not belonging to the European Union, both units being independent from each other: the UCI country code is allocated to the relevant residual extra-EU code.
Description of the treatment of multiple minority ownership
No particular treatment for this case.
3.5. Statistical unit
The statistical unit of FATS is the enterprise as defined in line with the Regulation (EEC) No 696/93 on the statistical units for the observation and analysis of the production system in the Community.
3.6. Statistical population
For all variables except for variables 230101 (Intramural R & D expenditure), 230301 (Intramural R & D expenditure in foreign-controlled enterprises), 230201 (R & D personnel) and 230401 (R & D personnel in foreign-controlled enterprises): Market producers of NACE Sections B to N and P to R and divisions S95 and S96;
For variables 230101 (Intramural R & D expenditure), 230301 (Intramural R & D expenditure in foreign-controlled enterprises), 230201 (R & D personnel) and 230401 (R & D personnel in foreign-controlled enterprises): Market producers of NACE Sections B to F.
3.7. Reference area
Luxembourg
3.8. Coverage - Time
2009-2023
3.9. Base period
Not applicable
Number of enterprises and employment variables are recorded in absolute figures.
Monetary data of enterprises are recorded in thousands EUR.
Data refers to the calendar year or, if available, the financial year.
6.1. Institutional Mandate - legal acts and other agreements
We apply the (n,k)-dominance rule, i.e. a cell is suppressed if n units separately or jointly dominate the total value of a cell by at least k%. The (n, k) parameters for Luxembourg are confidential. For any cells that are left after applying the sensitivity rule, a minimum frequency is applied. A cell is suppressed if there are less than n units in a given cell. The n parameter for Luxembourg is confidential.
Secondary confidentiality rules
The secondary suppression is calculated by tau-Argus using the ‘Modular’ algorithm. Manual suppressions or cost adjustments are performed to adjust the secondary confidentiality pattern calculated by the software.
a) Secondary suppression within a table
A cell is suppressed for secondary confidentiality if n units jointly or separately dominate the confidential subtotal by at least k%;
special attention is paid to the impact of singletons, a risk which is in most cases directly addressed by the tau-Argus Modular algorithm;
tau-Argus is set to minimise the cost when determining the secondary suppressed cells.
However, we also want to provide the user with relevant data, whether it is in terms of interpretation and/or availability of time series. Consequently, the cost minimisation can be overridden for economic and/or historical reasons.
b) Secondary suppression due to linked-table disclosure risks
A link is defined to exist between a cell sharing the same cell coordinates in two tables if an estimate for that cell based on the source table can be produced within p% range of the primary confidential cell's value of the target table. Most often, estimates based on the rule of three and linear interpolation, both of which are common user scenarios, are tested. Please note that p% only refers to the relationship between the dominance and p% thresholds and not to the p% sensitivity rule.
The following linked-table risks are addressed:
historical disclosure (time dimension): no primary historically confidential cell should be compromised by disclosing the same cell for the current reference year. As long as there is a significant link with a prior year primary confidential data, a cell may not be disclosed for the current reference year.
links to any other table sharing at least one dimension, including SBS tables by activty.
Other SDC policy considerations
The statistical confidentiality analyses are performed on the basis of turnover (shadow variable approach). The same pattern is therefore applied to all variables, including the number of enterprises.
For R&D expenditure and R&D personnel, a specific pattern is calculated based on R&D expenditure with the constraint that the confidentiality flags are furthermore inherited from the turnover-based confidentiality pattern.
No national dissemination of national IFATS statistics.
Annual for all variables except for variables 230101 (Intramural R & D expenditure), 230301 (Intramural R & D expenditure in foreign-controlled enterprises), 230201 (R & D personnel) and 230401 (R & D personnel in foreign-controlled enterprises): Market producers of NACE Sections B to N and P to R and divisions S95 and S96;
Biennial (every odd-numbered year) for variables 230101 (Intramural R & D expenditure), 230301 (Intramural R & D expenditure in foreign-controlled enterprises), 230201 (R & D personnel) and 230401 (R & D personnel in foreign-controlled enterprises): Market producers of NACE Sections B to F.
Data on STATEC's platform are usually available in SDMX format. However, IFATS tables have not yet been disseminated at national level.
10.3.1. Data tables - consultations
Not available.
10.4. Dissemination format - microdata access
Any national micro-data access is governed by article 16 of the Law of 10 July 2011 on the organisation of the National Institute for Statistics and Economic Studies.
10.5. Dissemination format - other
Data are transmitted annually to Eurostat either to be used in European aggregates and for country comparisons.
Inward FATS in Luxembourg are handled as an additional dimension to SBS data. Consequently, the underlying populations are the same, as are the economic data. The ultimate country of control breakdown, which is added by IFATS, is based on a multitude of statistical and administrative data, including the EuroGroupsRegister (EGR). Full mirroring with SBS statistics comes nonetheless at a price. Because the aforementioned data sources such as the EGR do not cover the whole SBS population in Luxembourg, it is not possible to use only hard measures in the statistical production process. The resulting imprecisions are dealt with by imputation and by the use of the facilities offered by the country code classification. The economic footprint of the imprecisions are very low, despite looking more significant in terms of the number of businesses. The inward FATS data could be further improved if Luxembourg was granted access to the administrative register of effective beneficiaires. This would help to fill in the remaining gaps for smaller businesses. Finally, in the time series, breaks can happen due to SBS data (i.e. inherited breaks) but also due to historical errors in ultimate ownership information sources, including EGR. The latter, when they significant, are flagged accordingly at the ultimate country level.
12.1. Relevance - User Needs
Users of inward FATS are: European Commission services, international organisations, ministries, chambers of commerce, trade unions, journalists, researchers etc. User needs go beyond what data compilers can provide.
12.2. Relevance - User Satisfaction
No user satisfaction survey conducted in this statistical area.
The overall accuracy of the Inward FATS can be assessed as very good:
The economic data are directly derived from SBS, without any cut-off threshold, and of very good quality.
At the level of country of ultimate control, the UCI approach accounts for more than 70% of the enterprises, while the rest is covered by UCI proxies. The quality of the country breakdown is of good quality.
While bigger businesses are often captured by surveys and the EuroGroupsRegister (EGR), the EGR's coverage and regular survey coverage are typically insufficient for smaller entities. EGR barely covers a third of the active enterprises in Luxembourg's SBS population. Luxembourg fills in part of this gap with administrative business register data and specialised commercial data. A significant portion of businesses, known to be under foreign control as per the statistical business register (2023: 23.9%), is allocated to the residual foreign-control country code ("Extra-EU not allocated") as a proxy. Nevertheless, the economic footprint of this residual country code is very low, with less than 5% when expressed in terms of either employment or turnover in 2023.
A negligible portion of businesses (less than 3%) has to be allocated by deductive imputation under national control. The economic footprint of this imputation is very low, more specifically 1.2% of employment and 0.5% of turnover in 2023.
Please refer to subsections 13.1.1 to 13.1.4 for further details.
13.1.1. Use of residual geographic codes (Extra EU-27 not allocated, etc.)
The residual code "Extra-EU not allocated" is used for handling two cases:
an enterprise is controlled equally by at least two distinct extra-EU countries
an enterprise is known to be foreign-controlled based on information available in the statistical business register (which draws this information from the once-in-a-livetime survey on economic activities), however, the precise country of ultimate control is unknown - both conditions need to be met. In this case, the residual code reflects the incompleteness of the hard data sources, including the EuroGroupsRegister, used to compile IFATS.
The annexed table here below shows a share of enterprises allocated to this residual country code is 23.9%, while the share of all the other characteristics for this residual code is less than 4% (e.g. employment 3.7%, turnover 0.8%). Therefore, this country code mainly concerns small enterprises.
13.1.2. UCI Approach applied to identify the relevant population of reporting units
The UCI approach is applied in accordance with the EBS regulation.
For the reference year 2023, the UCI approach represents 71.7% of the number of enterprises, 73.5% of turnover and 80.1% of the number of persons employed.
The following concepts are part of this approach:
ultimate beneficial owner (UBO),
worldwide headquarters or head office (WHQ),
ultimate consolidating parent undertaking (UCPU).
Proxies used for cases for which the UCI approach could not be implemented:
parent company,
other direct or indirect majority owners.
For usage rates, please refer to 13.1.4.
13.1.3. Update date (or frequency of updates) of the information regarding the country of the UCI by the “source administration”
Annual.
13.1.4. Description of other method used to improve the accuracy of the UCI
We used several approaches, given that not all pieces of information have the same quality:
The UCI approach (through its proxies UBO, WHQ and UCPU) is used whenever the available information allowed for it - for 71.7% (2022: 72.6%) of all SBS units in 2023, the UCI information is available. As of 2013, administrative and commercial data sources have been finally available as a structured file, allowing to identify by far more units as UCIs than ever before. Inevitably, this has caused several major breaks in the IFATS series.
An alternative approach, which includes parent companies and other direct or indirect majority owners, is used when we observe some ownership or control information for which we are unable to conclude whether or not the UCI has been identified. In some cases this information corresponds to the first foreign owner, in other cases even a unit further higher up in the control chain (often an intermediate parent company). This approach is used as a proxy to the UCI for 25.5% of all SBS units in 2023 (2022: 24.7%).
For a minor share of units, i.e. less than 0.01% of all SBS units (2022: <0.01%), we have some information on the truncated group in Luxembourg, which is used as a proxy to the UCI. Most of these units are nationally controlled.
For the remaining share of units, please refer to sections 13.3.1. and 20.1.
13.2. Sampling error
There is no separate survey for inward FATS. IFATS data are compiled from existing surveys and administrative sources.
13.2.1. Sampling error - indicators
Not available.
13.3. Non-sampling error
Please refer to subsections 13.3.1 to 13.3.5.
13.3.1. Coverage error
The SBS population is used as the frame for inward FATS and as a source at the micro-data level for the vast majority economic indicators. The only exception to the latter are R&D indicators, which are obtained from the R&D survey micro-data.
13.3.1.1. Over-coverage - rate
0%
13.3.1.2. Common units - proportion
100%.
For the geographical breakdown, undercoverage initially accounts for 2.8% of the units in the population. To avoid any attrition on the economic information due to undercoverage, the geographical breakdown has been imputed to "national control" for those units. In practice, this imputation concerns only SMEs, most of which employ less than 5 persons. After this deductive imputation, the common units proportion is back to 100%.
Also refer to 18.1. for further details.
13.3.1.3. Misclassification errors
Data on the geographical breakdown as well as the NACE code are subject to misclassification errors in the time series.
NACE code errors are dealt with in SBS data production, while the geographical breakdown specific to inward FATS is dealt with during IFATS data production.
In case that a misclassification is observed only for the current reference year, data editing procedures are applied to correct the misclassification. Error detection is based on materiality thresholds.
If a historical misclassification is observed in the past data due to improved quality data for the current reference year and if the impact of this historical misclassification on the time series by geographical breakdown exceeds a given threshold, the relevant table cells are flagged with the flag "break in series". The flagging is only performed by country of control and not by both country of control and activity.
Likewise, any "break in series" flags in SBS (mainly NACE misclassifications) are ported over to the relevant cells in the inward FATS tables, but not for the intersection of both country of control and activity.
13.3.1.4. Under- and over-coverage problems
Refer to 13.3.1.2.
13.3.2. Measurement error
Not available.
13.3.3. Non response error
There is no separate survey for inward FATS. IFATS data are compiled from existing surveys and administrative sources.
IFATS statistics are calculated annually for reference year T.
Data collection takes place at t+18 months after the end of the reference period.
Data transmission to Eurostat takes place at t+20 months.
Data dissemination at national level at the moment.
14.1.1. Time lag - first result
Not relevant.
14.1.2. Time lag - final result
Not available.
14.2. Punctuality
Data for reference year 2023 were transmitted to Eurostat with a minor delay.
14.2.1. Punctuality - delivery and publication
Restricted from publication
15.1. Comparability - geographical
Not available.
15.1.1. Asymmetry for mirror flow statistics - coefficient
Restricted from publication
15.2. Comparability - over time
Please refer to the subsections 15.1 to 15.4
15.2.1. Length of comparable time series
Length of time series: 2009-2023
Length of comparable time series: 2013-2023
15.2.2. Reasons and differences in concepts and measurement methods for breaks in time series
During the revision of SBS 2005-2011 data, inward FATS data were also analysed for their consistency throughout the period 2009 - 2011. The first reason for non-comparability is the non-comparability of SBS data for the period 2009 to 2011. The second reason is the detection of measurement errors for some UCI countries (in hindsight due to more stable data). A revision of IFATS 2009-2011 data was not produced due to the lack of resources necessary to deal with the statistical disclosure control bottleneck.
Furthermore, a general break in series is observed as of 2013. This is due to higher quality data sources, more particularly the national enterprise group register, additional administrative and commercial sources. The break in series is bound to stay due to the lesser quality of the sources available before the reference year 2013.
In 2015, a break in series was recorded for 18 geographical codes. Some of the breaks were caused by significant errors in historical data (e.g. wrong UCI country, detection of group-only special purpose entities, which are off-scope in SBS / IFATS). However, quite a few breaks were caused by a decrease of the quality of the national enterprise group register 2015 data due to an apparent change in handling natural persons as global decision centres. Given the increased use of EGR in the said dataset (refer to section 18), it is likely that this issue could be related to the EGR 2.0 production process.
As of 2021, with the entry into force of the EBS regulation, the coverage of IFATS has been extended to sections K, P, Q, R. During the implementation of these changes, special care has been taken not to break with the time series before 2021.
The events concerning the reference years 2015 and 2021 did not cause any generalised break but only breaks at the "local" level.
15.3. Coherence - cross domain
Please refer to subsections 15.3.1 to 15.3.7.
15.3.1. Coherence - sub annual and annual statistics
Not applicable
15.3.2. Coherence - National Accounts
No such analysis performed. On a methodological level, national accounts by institutional sector in Luxembourg tend to prefer the first foreign counterpart approach (FFC) over the ultimate controlling institutional unit approach (UCI). In addition, unlike IFATS, national accounts are not limited to the business market economy. Therefore, it is not possible to perform a useful comparison.
15.3.3. Coherence – National Statistical Business Register (NSBR)
Not an issue.
15.3.4. Coherence – Structural Business Statistics (SBS)
No incoherence because they are treated as the same population.
15.3.5. Coherence – R & D
No incoherence.
15.3.6. Coherence – Foreign Direct Investment (FDI)
No such analysis performed.
15.3.7. Coherence – EuroGroups Register (EGR)
EGR coverage is calculated in the framework of the LU EGR metadata reporting. This coverage assessment should be interpreted with care:
1) Units having the residual country code D09 are usually removed from the denominator, leading to a significantly overstated coverage proportion. When adjusting for the residual country code D09, EGR covers only 70% of the number of foreign-controlled enterprises in the SBS population in 2023, meaning that for 30% of the enterprises, the foreign-country of control information depends on other data sources.
2) By design, EGR covers neither natural persons nor smaller independent legal persons. As a matter of fact, EGR only covers 29% of the LU total SBS population in 2023, that is including both national-controlled and foreign-controlled businesses, meaning that for 71% of the enterprises, the country of control information depends on data sources other than EGR.
15.4. Coherence - internal
Internal inconsistencies stem most often from rounding errors.
There is no separate survey for inward FATS. IFATS data are compiled from existing surveys and administrative sources.
Any data integration costs are borne exclusively by the NSI and do not impact any respondent within any of those data sources.
For 97.2% of the total SBS population in 2023, information on the UCI or a proxy was available, the share of dependence (in terms of number of units) by source (including cold-deck imputations for each source) being as follows:
Business Register: 43.7% (2022: 41.8%);
Register of commerce: 40.8% (2022: 43.1%);
EGR: 6.4% (2022: 5.4%);
BVD: 6.3% (2022: 6.8%);
SBS: 1.0% (2022: 1.1%);
FDI: <0.1% (2022: <0.1%);
Other sources: 1.8% (2022: 1.7%).
Slightly less than half of those data were made available through the national enterprise group register.
For the big enterprises (at least 250 persons employed), the share of dependence on the Business Register as a source is 6.3% in 2023 (2022: 8.0%), the other sources being FDI (1.3%, 2022: 1.8%), BVD (17.3%, 2022: 16.1%), RCS (24.9%, 2022: 24.1%), and SBS (4.6%, 2022: 4.5%) as well as some expert assessments for which the source is unknown (2023: 21.5%). Data from EGR accounted for 20.3% (2022: 17.9%). The national enterprise group register centralised more than half of these sources.
For 2.8% of the total SBS population the assumption was made that Luxembourg was the country of control (“LU control assumption”, 2022: 2.8%). This assumption was only made when there was no other information available – basically, this assumption exists only for SMEs, most of which employ less than 5 persons. Given that the vast majority of employment and turnover of the total SBS population were covered through UCI information or proxies, we deemed it reasonable to make use of this assumption. We decided not to employ any survey estimation methods, given that the subset of SMEs for which information was available is not the result of a random sampling method (risk of selection bias) and because data available for grossing-up per strata was scarce. We decided not to apply any cut-off threshold as this would have created inconsistencies with the SBS tables.
18.1.1. Methodological approach
Inwards FATS data are obtained by combining SBS and R&D (economic information) and any survey data and administrative sources for the geographical breakdown at the micro-data level. SBS is furthermore used to provide the population for inward FATS.
As of the reference year 2013, the preferred data source is the national variety of the enterprise group register. This register includes the data from EuroGroupsRegister where consistent with data available at the national level. The value-added of the national register instance is the coverage of SMEs as well as the increased consistency with FDI. For units not covered through the national enterprise group register, data is obtained either from the SBS survey, a one-shot economic activity survey or approximated through the business register (e.g. foreign or national control via the institutional sector code, address for the natural persons, proxy based on legal form, etc.).
After data integration, the output micro-data and tabular data are checked for coherence in time and across applicable domains. Exceptions to these tests are analysed, documented, validated and, if necessary, edited using a working area.
For usage rates of survey data and administrative sources pertaining to the concept of control, please refer to section 18.1.
18.1.2. Use of cut-off thresholds
No cut-off threshold used.
18.2. Frequency of data collection
Annual.
18.3. Data collection
Source to define your population:
0 % census survey
0 % sample survey
100 % Structural Business Statistics (SBS)
0 % Foreign Direct Investment (FDI)
0 % EuroGroups Register (EGR)
0 % Statistical Business Register (SBR)
0 % Administrative sources
0 % Private data sources
0 % Publicly available sources
0 % Other data sources (indication of the sources: ...)
Economic data obtained using:
0 % census survey
0 % sample survey
100 % Structural Business Statistics (SBS)
0 % Foreign Direct Investment (FDI)
0 % EuroGroups Register (EGR)
0 % Statistical Business Register (SBR)
0 % Administrative sources
0 % Private data sources
0 % Publicly available sources
0 % Other data sources (indication of the sources: ...)
18.4. Data validation
After data integration, the output micro-data and tabular data are checked for coherence in time and across applicable domains.
Exceptions to these tests are analysed, documented, validated and, if necessary, edited using a working area.
The following tests are deployed:
* consistency in time at the micro-data level based on a materiality threshold and for period of two and three consecutive reference years,
* consistency with outward FATS UCIs at the micro-data level,
* consistency with the national enterprise group register is verified at the micro-data level since reference year 2013,
* consistency in time for tabular data by country and by activity (2-digit level) based on a materiality threshold.
18.5. Data compilation
Please refer to the subsections 18.5.1 to 18.5.3.
18.5.1. Imputation - rate
2.8% of the number of units for the geographical breakdown.
18.5.2. Use of a method to deal with non-response (both unit and item non-response)
No imputation of economic information because all indicators are drawn from SBS and business R&D micro-data.
Missing geographical information is dealt with by deductive imputation to avoid undercoverage.
18.5.3. Share of estimated values
2.8%
18.6. Adjustment
Not applicable.
18.6.1. Seasonal adjustment
Not applicable.
None.
Foreign Affiliates Statistics (FATS) measure the commercial presence through affiliates in foreign markets.
Inward foreign affiliates statistics (IFATS) shall mean statistics describing the activity of foreign affiliates resident in the compiling country.
In country-level business statistics foreign-controlled enterprise shall mean an enterprise resident in the compiling country over which an ultimate controlling institutional unit not resident in the compiling country has control.(Table 14 of the EBS Implementing Regulation (EU) 2020/1197).
Variables on the country-level business activities in the IFATS data category:
Business activities in foreign control:
210301. Number of foreign-controlled enterprises
220501. Number of employees and self-employed persons in foreign-controlled enterprises
220701. Employee benefits expense in foreign-controlled enterprises
230301. Intramural R & D expenditure in foreign-controlled enterprises
230401. R & D personnel in foreign-controlled enterprises
240301. Total purchases of goods and services of foreign-controlled enterprises
240302. Purchases of goods and services for resale of foreign-controlled enterprises
250601. Net turnover of foreign-controlled enterprises
250701. Value of output of foreign-controlled enterprises
260201. Foreign-controlled enterprises’ gross investment in tangible non-current assets
250801. Value added of foreign-controlled enterprises
Business activities in total economy:
210101. Number of active enterprises
220101. Number of employees and self-employed persons
220301. Employee benefits expense
230101. Intramural R & D expenditure
230201. R & D personnel
240101. Total purchases of goods and services
240102. Purchases of goods and services for resale
250101. Net turnover
250301. Value of output
250401. Value added
260101. Gross investment in tangible non-current assets
27 October 2025
Foreign Affiliates Statistics (FATS) measure the commercial presence through affiliates in foreign markets.
Inward foreign affiliates statistics (IFATS) shall mean statistics describing the activity of foreign affiliates resident in the compiling country.
In country-level business statistics foreign-controlled enterprise shall mean an enterprise resident in the compiling country over which an ultimate controlling institutional unit not resident in the compiling country has control.(Table 14 of the Implementing Regulation (EU) 2020/1197).
Foreign affiliate in the framework of outward FATS is an enterprise or branch not resident in the compiling country over which an institutional unit resident in the compiling country has ultimate (direct or indirect) control.
Domestic affiliate shall mean an enterprise resident in the compiling country over which a UCI resident in the same compiling country has control.
Ultimate Controlling Institutional of a foreign affiliate (UCI) shall mean the institutional unit, proceeding up a foreign affiliate’s chain of control, which is not controlled by another institutional unit.
Control is the ability to determine the general policy of the affiliate by choosing appropriate directors, if necessary. In this context, enterprise A is deemed to be controlled by an institutional unit B when B controls, whether directly or indirectly, more than half of the shareholders' voting power or more than half of the shares.
Indirect control means that an institutional unit may have control through another affiliate which has control over enterprise A.
Active enterprise is a statistical units which at any time during the reference period was ‘enterprise’, as defined in Regulation (EEC) No 696/93, and also active during the same reference period. A statistical unit is considered to have been active during the reference period if, in said period, it either realized positive net turnover or produced outputs or had employees or performed investments.
Employees and self-employed persons are persons who work for an observation unit on the basis of a contract of employment and receives compensation in the form of wages, salaries, fees, gratuities, piecework pay or remuneration in kind; and persons who are the sole owners or joint owners of the statistical unit in which they work. Family workers and outworkers, whose income is a function of the value of the outputs of the statistical unit, are also included.
Employee benefits expense contains all expenses arising in relation with employee benefits, recognized by the statistical unit during the reference period. Those are are all forms of consideration given by the statistical unit in exchange for service rendered by employees or for the termination of employment.
Research and experimental development (R & D) comprise creative and systematic work undertaken in order to increase the stock of knowledge – including knowledge of humankind, culture and society – and to devise new applications of available knowledge. Expenditures on intramural R & D represent the amount of money spent on R & D that is performed within a reporting unit. Intramural R & D expenditures are all current expenditures plus gross fixed capital expenditures for R & D performed within a statistical unit during a specific reference period whatever the source of funds. R & D current expenditures include labour costs for internal R & D personnel and other current costs (costs for external R & D personnel, purchase of services.). Gross fixed capital expenditures for R & D include: acquisition of land, acquisition of buildings, acquisition of information and communication equipment, acquisition of transport equipment, acquisition of other machinery and equipment, acquisition of capitalised computer software, acquisition of other intellectual property products.
R & D personnel in a statistical unit include all persons engaged directly in R & D, whether employed by the statistical unit or external contributors fully integrated into the statistical unit’s R & D activities, as well as those providing direct services for the R & D activities (such as R & D managers, administrators, technicians and clerical staff).
Total purchases of goods and services contains all amount of goods and services purchased by the statistical unit, recognized in accounting as either current assets or expenses during the reference period.
Purchases of goods and services for resale in are purchases of goods for resale to third parties without further processing. It also includes purchases of services by ‘invoicing’ service companies, i.e. those whose turnover is composed not only of agency fees charged on a service transaction (as in the case of estate agents) but also the actual amount involved in the service transaction, e.g. transport purchases by travel agents.
Net turnover consists of all income arising during the reference period in the course of ordinary activities of the statistical unit, and is presented net of all price reductions, discounts and rebates granted by it.
Value of output represents the value of the total output of the statistical unit, generated during the reference period.
Value added is a composite indicator of net operating income, adjusted for depreciation, amortization and employee benefits, all components being recognized as such by the statistical unit during the reference period.
Gross investment in tangible non-current assets includes all additions to tangible non-current assets, recognized as such by the statistical unit during the reference period, except any increases from revaluations or reversals of previously recognized impairment losses and from reclassifications (transfers) of other tangible non-current assets.
Description of the coverage of Special Purpose Entities (SPE)
SPEs are dealt with in three ways:
If the SPE is a "resident" unit under control, the SPE is dealt with during SBS data production. Any SPE present or absent in the SBS population is present or absent in the IFATS population;
Luxembourgish SPEs appointed as ultimate controlling institutional unit (UCI) are subject to validation, if they can be detected based on the thresholds or if they are classified under NACE code K 64.2 or 64.9. By definition, SPEs have little substance in the host economy and are therefore rarely appointed as controlling units. The definition of control suggests sufficient economic substance in the host economy. Control is also required to be active (i.e. local) rather than passive (i.e. remote-controlled).
Foreign SPEs appointed as UCI are accepted as such, unless we have sufficient data to prove that the UCI should be in a different country. UCIs assigned to offshore countries may hint at such SPEs.
Description of the treatment of equally shared control
We distinguish 5 types of equally shared control, for which we apply the geographical proximity principles:
control shared between one national (LU) and one foreign unit which are not linked by an enterprise group (i.e. both units being independent from each other): in this case we assign ultimate control to the the resident unit itself, the UCI country thus being under national control;
control shared between two independent units resident in the same country: the UCI country code is allocated to the said country of residence, no matter whether the country is LU (national), EU or extra-EU;
control shared between one country belonging to the European Union and one non-EU country, both units being independent from each other: the resident unit is under equally shared foreign control. Given the presence of a EU country, the UCI country code is allocated to the country of residence in the EU (dominant controller by assumption);
control shared between two different countries belonging to the European Union, both units being independent from each other: the resident unit is under joint foreign control, therefore the UCI country code is allocated to the relevant residual EU code;
control shared between two different countries not belonging to the European Union, both units being independent from each other: the UCI country code is allocated to the relevant residual extra-EU code.
Description of the treatment of multiple minority ownership
No particular treatment for this case.
The statistical unit of FATS is the enterprise as defined in line with the Regulation (EEC) No 696/93 on the statistical units for the observation and analysis of the production system in the Community.
For all variables except for variables 230101 (Intramural R & D expenditure), 230301 (Intramural R & D expenditure in foreign-controlled enterprises), 230201 (R & D personnel) and 230401 (R & D personnel in foreign-controlled enterprises): Market producers of NACE Sections B to N and P to R and divisions S95 and S96;
For variables 230101 (Intramural R & D expenditure), 230301 (Intramural R & D expenditure in foreign-controlled enterprises), 230201 (R & D personnel) and 230401 (R & D personnel in foreign-controlled enterprises): Market producers of NACE Sections B to F.
Luxembourg
Data refers to the calendar year or, if available, the financial year.
The overall accuracy of the Inward FATS can be assessed as very good:
The economic data are directly derived from SBS, without any cut-off threshold, and of very good quality.
At the level of country of ultimate control, the UCI approach accounts for more than 70% of the enterprises, while the rest is covered by UCI proxies. The quality of the country breakdown is of good quality.
While bigger businesses are often captured by surveys and the EuroGroupsRegister (EGR), the EGR's coverage and regular survey coverage are typically insufficient for smaller entities. EGR barely covers a third of the active enterprises in Luxembourg's SBS population. Luxembourg fills in part of this gap with administrative business register data and specialised commercial data. A significant portion of businesses, known to be under foreign control as per the statistical business register (2023: 23.9%), is allocated to the residual foreign-control country code ("Extra-EU not allocated") as a proxy. Nevertheless, the economic footprint of this residual country code is very low, with less than 5% when expressed in terms of either employment or turnover in 2023.
A negligible portion of businesses (less than 3%) has to be allocated by deductive imputation under national control. The economic footprint of this imputation is very low, more specifically 1.2% of employment and 0.5% of turnover in 2023.
Please refer to subsections 13.1.1 to 13.1.4 for further details.
Number of enterprises and employment variables are recorded in absolute figures.
Monetary data of enterprises are recorded in thousands EUR.
Please refer to the subsections 18.5.1 to 18.5.3.
For 97.2% of the total SBS population in 2023, information on the UCI or a proxy was available, the share of dependence (in terms of number of units) by source (including cold-deck imputations for each source) being as follows:
Business Register: 43.7% (2022: 41.8%);
Register of commerce: 40.8% (2022: 43.1%);
EGR: 6.4% (2022: 5.4%);
BVD: 6.3% (2022: 6.8%);
SBS: 1.0% (2022: 1.1%);
FDI: <0.1% (2022: <0.1%);
Other sources: 1.8% (2022: 1.7%).
Slightly less than half of those data were made available through the national enterprise group register.
For the big enterprises (at least 250 persons employed), the share of dependence on the Business Register as a source is 6.3% in 2023 (2022: 8.0%), the other sources being FDI (1.3%, 2022: 1.8%), BVD (17.3%, 2022: 16.1%), RCS (24.9%, 2022: 24.1%), and SBS (4.6%, 2022: 4.5%) as well as some expert assessments for which the source is unknown (2023: 21.5%). Data from EGR accounted for 20.3% (2022: 17.9%). The national enterprise group register centralised more than half of these sources.
For 2.8% of the total SBS population the assumption was made that Luxembourg was the country of control (“LU control assumption”, 2022: 2.8%). This assumption was only made when there was no other information available – basically, this assumption exists only for SMEs, most of which employ less than 5 persons. Given that the vast majority of employment and turnover of the total SBS population were covered through UCI information or proxies, we deemed it reasonable to make use of this assumption. We decided not to employ any survey estimation methods, given that the subset of SMEs for which information was available is not the result of a random sampling method (risk of selection bias) and because data available for grossing-up per strata was scarce. We decided not to apply any cut-off threshold as this would have created inconsistencies with the SBS tables.
Annual for all variables except for variables 230101 (Intramural R & D expenditure), 230301 (Intramural R & D expenditure in foreign-controlled enterprises), 230201 (R & D personnel) and 230401 (R & D personnel in foreign-controlled enterprises): Market producers of NACE Sections B to N and P to R and divisions S95 and S96;
Biennial (every odd-numbered year) for variables 230101 (Intramural R & D expenditure), 230301 (Intramural R & D expenditure in foreign-controlled enterprises), 230201 (R & D personnel) and 230401 (R & D personnel in foreign-controlled enterprises): Market producers of NACE Sections B to F.
IFATS statistics are calculated annually for reference year T.
Data collection takes place at t+18 months after the end of the reference period.
Data transmission to Eurostat takes place at t+20 months.
Data dissemination at national level at the moment.
