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.
Structural business statistics (SBS) describes the structure, conduct and performance of economic activities, down to the most detailed activity level (several hundred economic sectors).
SBS covers all activities of the non-financial business economy with the exception of agricultural activities and personal services. Limited information is available on banking, insurance and pension funds.
Main characteristics (variables) of the SBS data category:
Business demographic variables (e.g. number of enterprises),
"Output related" variables (e.g. turnover, value added),
"Input related" variables: labour input (e.g. employment, hours worked); goods and services input (e.g. total of purchases); capital input (e.g. material investments).
3.2. Classification system
Statistical Classification of Economic Activities in the European Community (NACE): NACE Rev.1 was used until 2001, NACE Rev. 1.1 since 2002, and NACE Rev 2 is used from 2008 onwards. Key data were double reported in NACE Rev.1.1 and NACE Rev.2 for 2008. From 2009 onwards, only NACE Rev.2 data are available.
The regional breakdown of the EU Member States is based on the Nomenclature of Territorial Units for Statistics (NUTS). Detailed information about the consecutive NUTS Regulations can be found at Eurostat's website.
3.3. Coverage - sector
The SBS coverage was limited to Sections C to K of NACE Rev.1.1 until 2007. Starting from the reference year 2008 data is available for Sections B to N and Division S95 of NACE Rev.2. With 2013 as the first reference year information is published on NACE codes K6411, K6419 and K65 and its breakdown.
Reporting and observation unit: legal and enterprise unit.
From the reference year 2018 the scope of statistics broadened with a new observation unit: the statistical unit enterprise. Monitoring and consolidating the internal flow between individual members of the enterprise makes it possible to measure the process of the national value creation more accurately.
There are three groups of enterprises:
SIMPLE ENTERPRISES: The legal entity is not related to another company, i.e. an enterprise unit can be matched to a legal entity.
COMPLEX SMALL ENTERPRISES: for the enterprises created by linking 2-3 legal units, an automatic consolidation algorithm has been developed in accordance with international recommendations.
COMPLEX LARGE ENTERPRISES: for enterprises with a minimum of 4 legal units, a questionnaire is used to observe each member of the enterprise and to compile data for internal flow. For this purpose the questionnaire 2414 (‘Survey on Enterprises’ - enterprises with at least 4 legal units, where the aggregate turnover of the legal units in the previous period reaches HUF 500 million) has been developed.
3.6. Statistical population
SBS population is identified on the base of the Business Register. The Business Register provides the frame of the SBS: the legal units /enterprise statistical units for which statistical survey (performance, labour, investment) or administrative data are available, will be involved in SBS population. None of the NACE activity codes or size classes are excluded from the target population.
3.7. Reference area
SBS data covers the territory of Hungary. All regions are covered by the SBS data.
SBS reference population includes foreign branches of legal units registered in the territory of Hungary which do not constitute a separate legal entity and are dependent on the resident legal unit.
Branches of foreign legal units are involved in SBS as well, while the performance (output) related activities abroad of Hungarian legal units (foreign branches of resident units which constitute a separate legal entity) are excluded according to the Regulation (EC) No 295/2008.
Accordingly, Hungarian SBS data is consistent with ESA2010 and BPM6 international regulations requirements.
3.8. Coverage - Time
SBS data are available for years: 2001-2020.
In 2008 the Nace Rev. 2 has been introduced, so the time-series broken down by activity are not comparable with the former data series. Length of comparable time series: 2008-2017. There is a break in the series in 2018 due to the change of the reporting and observation unit. Until 2018, the HCSO reported data at the level of legal units.
3.9. Base period
Not applicable.
Number of enterprises and number of local units are expressed in units.
Monetary data are expressed in millions of €.
Employment variables are expressed in units.
Per head values are expressed in thousands of € per head.
Ratios are expressed in percentages.
2020
SBS data refers to calendar year. Units are known applying a fiscal year different from the calendar year are surveyed (regardless of the size-class).
6.1. Institutional Mandate - legal acts and other agreements
Government Decree 184/2017 (VII. 5.) on implementing Act CLV of 2016 on Official Statistics
Act CXII of 2011 on Informational self-administration and freedom of information.
The confidentiality policy of HCSO is available on its website.
7.2. Confidentiality - data treatment
HCSO disseminate only aggregated data in full compliance with the rules of confidentiality. Individual data, as well as aggregated data consisting of fewer than 3 legal units are regarded as confidential and therefore not published. Researchers have access to de-identified data sets and to anonymised micro data for scientific purposes with appropriate legal and methodological guaranties in place. As for the employees, they can work with datasets in their competence with registered and controlled access rights. For details see the information on confidentiality for data providers on the HCSO website.
Rules applied for de-identifying the confidential data in SBS:
primary confidentiality (A cell is flagged as confidential if less than three units ('too few units') contribute to the total of that cell)
secondary confidentiality (Treating a non-confidential cell as confidential in order to prevent disclosure of a confidential cell)
8.1. Release calendar
Dissemination calendar contains the release dates of SBS datasets involved in the Dissemination database. The release calendar on the website of HCSO provides information not only about the first releases, STADAT and dissemination tables but the analyses, methodological publications, promotional publications, reports, statistical reflections, yearbooks and pocketbooks.
All users can have equal access to statistical information at the same time, pre-release access is exceptional, its rules are public, and their violation – by external or internal users or persons – has legal implications.
Data that are not published on the website of HCSO can be disseminated as well. There are two main types of disclosure depending on the requiring:
Public authorities: data transfer is obligatory under a specific law (in this case the data supply is free of charge) or based upon a signed contract for data transfer agreement;
Individual requests (companies, researchers etc.)
(HCSO supports scientific researches with the available tools, by extending the scope of data usable for research purposes and data access possibilities.)
Payments need to be made for information which is not available as a result of standardized work processes, so they need to be compiled upon individual requests, and which generates additional personnel and material costs.
Annual
10.1. Dissemination format - News release
Not applicable.
10.2. Dissemination format - Publications
SME publications (based on SBS data) are available in HCSO’s publication repertory in Hungarian language.
Dissemination database tables are published in the middle of November, 10,5 months after the reference period (in case of preliminary data), while the second publishing (final data) is by the end of August (which means 20 months) after the reference year.
Availability:
HCSO’s website: Data, publications - Dissemination database - Economic Statistics – Business Statistics - Annual performances and expenses of enterprises / Indicators of annual performance of enterprises
10.4. Dissemination format - microdata access
SBS data are not published at micro-level and the micro-data access is not an available option as the data consists of survey and administrative data as well. The data cannot be released at individual level.
HCSO provides anonymised microdata sets for the researcher for scientific purposes. This activity is based on contract between the research institute and the Statistical Office. The access is granted only for approved research project that meet all researcher accreditation criteria. For detailed information see access to anonymised microdata sets on the website of HCSO.
For the purpose of scientific research of statistical data, a Research room has been created where the access to anonymised microdata is provided in strict compliance with the high-level protection of individual data and the rules of data protection. Services of the Research room are available only after the Research room contract and the Confidentiality declaration were signed.
Research is possible with remote access as well.
In this case, the researcher can access micro-data in a secure IT environment from specified access points in fully the same operating conditions as the operation environment of the Research room. As in the environment of the Researcher room, the researcher receives the results of the research which are verified in the aspect of the risk of disclosure and not the micro-data forming the basis of the research themselves. The rules of using the secure environment and the terms of the use of results are determined by a contract between the researcher and the HCSO.
Release of anonymised microdata sets (similar to scientific use files), where direct identifiers are removed and further statistical disclosure control methods have been applied, as well as access to de-identified, but more detailed data in the onsite safe centre, through secure remote access and remote routine execution, are reserved to researchers only. Anonymised data sets are customised, and only the variables requested are provided to the researcher.
All data is checked for statistical disclosure before dissemination, in order to reduce the risk of disclosure of the statistical units to a possible minimum. When fulfilling individual requests for access to data, various tools of methodological, physical and legal protection are being applied depending on the nature of the request. Additional information may be found on statistical disclosure control methods among the rules for access to statistical data.
SBS data (performance, labour and investment data) are involved in the Dissemination database. The database can be found on the HCSO's website. Users can compile cross tables by choosing the indicators, period of time, size class or region.
STADAT (the acronym stands for statistical data) contains primarily time series. Performance data by size class and by SME categories based on the SBS data are involved in STADAT. STADAT tables are static tables about numerous topics, with methodological notes.
Yearly SME publication (using SBS data) is available at the HCSO’s website in Hungarian language.
SBS data files are transmitted to Eurostat according to the requirements of Regulation (EC) No 250/2009 and No 251/2009 of the European Parliament and of the Council of 11 March 2009 implementing and amending Regulation (EC) No 295/2008 of the European Parliament and of the Council.
10.6. Documentation on methodology
There are methodological notes (Metainformation) on HCSO's website about concepts and description of the statistical domain. The methodological notes are available in Hungarian and in English. Additional methodological comments can be found in Dissemination database (by clicking on the question mark icon next to the name of the given variable).
10.7. Quality management - documentation
There is a public documentation on HCSO's website which contains information about the main characteristics of data quality. The documentation is available in Hungarian and in English.
11.1. Quality assurance
European Statistics Code of Practice is on the website of HCSO with the Peer review on the implementation of European Statistics Code of Practice. The compliance with the Code of Practice is monitored regularly.
There are some additional documents in English on the website of HCSO:
The international legal acts (SBS Regulation and Commission Regulation No. 250/2009 and No. 251/2009 of 11 March 2009) prescribe the reference population, the definition of characteristics, the level of the different breakdowns (activity, size class, region, environmental domains, product) and the technical format of data sending.
The data production and transmission has been developed and all variables are published in line with these regulations.
HCSO is committed to permanent quality development and ensuring the credibility of statistics.
The day of the publication of data is specified in dissemination calendar and the deadlines are always met. The data transmission to Eurostat is always completed within deadlines.
The wide range of SBS data is available on the website of HCSO for everybody. There are not known serious quality problems.
12.1. Relevance - User Needs
Principal internal user: National Accounts Department
Relevant external users:
In case of SBS data, the most important foreign user is Eurostat, while at national level the main users are the ministries and other governmental organizations. These institutions usually specify their requests in special data transfer agreements or make it mandatory by law.
Additional needs cover mainly breakdowns different from those published (e.g. county level breakdown instead of regional) and these especially refer to the SME data.
SBS concepts and methods are based on European legislation.
SBS data available on the website of HCSO are not different from the data published by Eurostat, which means that the methodological content of variables, the classifications, the levels of the breakdowns are the same. Nevertheless, HCSO’s database does not contain all SBS variables which are available in Eurostat’s database. Environmental variables are unpublished as they belong to another statistical domain (i.e. Environmental statistics and accounts). It should be noted that the lack of these characteristics does not cause any problem as our users do not need these data.
In addition, the database of HCSO differs from the Eurostat database in that the data are published both at legal unit and enterprise level, while Eurostat only publishes data at enterprise level.
12.2. Relevance - User Satisfaction
In order to meet the users’ requests more, the Dissemination Database of HCSO is continuously developed (considering the number of confidential cells vs the amount of available data). There are regular (written) negotiations with the experts of ministries – in particular the Ministry of Innovation and Technology and the Ministry of Finance – to assess the emerging new needs.
Additional needs cover mainly breakdowns different from those published (e.g. county level breakdown instead of regional) and these especially refer to the SME data.
Number of downloads of 2020 year SBS data at legal unit level: 7.313 downloads
Number of downloads of 2020 year SBS data at enterprise level: 3.305 downloads
12.3. Completeness
Eurostat requirements are fulfilled (no missing variables or breakdowns).
13.1. Accuracy - overall
Sampling, coverage and measurement errors: not applicable
Only the non-response rate (0.82 percent in case of final SBS data at legal unit level) has a limited effect on the overall accuracy.
Preliminary data are published on 15th of November (in year t+1), while the final data are released in the end of August (in year t+2).
13.2. Sampling error
Sampling error is negligible.
Sample used in size categories 0-4 and 5-9 are not functioning as real samples anymore (in the very beginning when the administrative data sources were unavailable samples were used). It means that there is no grossing up, the samples are used for estimation purposes only (in the lower size categories only administrative data sources are available but not all of them is sufficiently detailed so estimations are done, e.g. in the case of ‘Tax return on non-flat rate tax of sole proprietors’ where the amount of material cost and the services come together and to define the above mentioned variables separately we have to rely on estimates).
Sampling error is not calculated as there is no grossing up from the sample data. Administrative data source is used for 'mass imputation'.
13.3. Non-sampling error
Characteristics used for weighting non-response: number of persons employed
Evaluation of the recorded unit non-response rate: very low
Non-sampling error is negligible.
In case of legal units with less than 20 persons employed there are neither estimations nor imputations for the variables 13131, 13411, 15210 and 16150.
The recorded unit non-response rate is lower than the rate for the reference year 2019. The data of non-response units (0.82% at legal unit level) are imputed on micro level on the basis of previous year's data. This way the bias arising from non-response is almost zero.
Coverage error is almost none.
There are no out-of scope units.
14.1. Timeliness
Dissemination database tables (both at legal unit and enterprise level) are published 10,5 months after the reference period (in case of preliminary data), while the second publishing (definitive data) is by the end of August (which means 20 months after the reference year).
14.2. Punctuality
The disseminations/publications are in line with the dates of dissemination calendar.
15.1. Comparability - geographical
The whole country is treated in a uniform manner from a statistical point of view. The same statistical concepts are applied for the entire territory of Hungary. There are no difficulties concerning the geographical comparability.
15.2. Comparability - over time
2008-2017
2018-2020
In 2008 the Nace Rev. 2 has been introduced, so the time-series broken down by activity are not comparable with the former data series.
There is a break in the series in 2018 due to the change of the reporting and observation unit. Until 2018, the HCSO reported data at the level of legal units.
15.3. Coherence - cross domain
SBS data are coherent with National Accounts and Prodcom statistics.
Data for SBS and National Accounts purposes are stored and produced in one common database (Joint Production Database). The database contains elementary level data and is the initial database of SBS series and the GVA by ESA. When creating the performance surveys (annual and intra-annual), during the elaboration of methodology and the data validation the experts of SBS and NA strongly cooperate. Existing differences between these two areas are caused by methodological discrepancies between SBS and ESA 2010 regulation (e.g. ESA follows the institutional sector classification by ESA2010 and there are additional differences due to the treatment of taxes and subsidies on production in case of some variables). Replacing SBS regulation with EBS will reduce the methodological divergences.
Three main areas of SBS indicators are: performance, labour and investment statistics.
All of these data are covered by the tax return of double-entry bookkeepers and annual financial report (profit and loss account, balance sheet, notes) of companies which are the most important basis of data checking and validation. In case of investment data building diaries, financial assistance for tangible investments granted by the EU are also applied for validation purposes.
Every annual indicator included in SBS is compared to its corresponding contained in intra-annual surveys (quarterly and monthly survey on performance statistics are checked with annual performance survey; quarterly survey on investments statistics are compared to annual investment survey and monthly labour survey are verified to the annual labour survey).
Consistency checks are performed between the following SBS characteristics and characteristics with similar methodological content of surveys/statistics.
These are the following:
‘Turnover’ and ‘value of export revenue’: Prodcom statistics (NSDCP) 1039 ‘Annual survey of industrial products and business services’ involved in SBS Annex VIII (NSDCP) 2146 ‘Sales structure of business services’ are checked to the data of Annex I-IV.
Not just the ‘total turnover’, but the ‘turnover by activity breakdown’ according to NACE Rev. 2 (separate table within (NSDCP) 2239 ‘Annual performance survey’) is compared to the product level (Prodcom) data.
‘Wages and salaries’, ‘number of employees’ are checked to the (NSDCP) 1117 ‘Labour cost survey’.
‘Personnel cost’ and ‘number of persons employed’ are compared to (NSDCP) 1117 ‘Labour cost survey’.
‘Wages and salaries’ are compared to the data of regional surveys (NSDCP) 1846 ‘Survey of local units’, covers NACE sections: B-F, H, J-N, S; (NSDCP) 1852 ‘Survey of local units of retail trade and catering services’, covers NACE sections G and I.
SBS’ ‘Gross investment in tangible goods data’ (NSDCP) 2240 ‘Annual investment survey’ are compared to (NSDCP) 1933 ‘Annual survey of investment structure’.
We run several plausibility checks on the (Annual Statistical Survey on Enterprises) ASSE data sent to us (NSDCP) 2414: a) To the total data of the legal units (the internal flow cannot be larger than the total flow of the legal unit) b) The reported data of the partners are also compared to each other.
All of the above mentioned data collections and statistics are synchronized at the stage of survey planning, which furthermore includes the coordinated elaboration (or refinement) of methodological descriptions and the preparation of consistency checks.
Surveys and their methodological descriptions include instructions for the data suppliers regarding the consistency rules that must be kept (internal consistency and rules of cross checks between surveys with similar data content).
These are the following:
(NSDCP) 2239 ‘Annual performance survey’ and (NSDCP) 2235, 2236 monthly and quarterly survey on performance statistics: The value of following variables must be equal: turnover, export revenue, capitalized production, purchases of goods and services (at the level of its components: material costs, goods and services purchased for resale in the same condition as received, contracted services), stocks (self-constructed assets: work in progress, intermediate, semi-finished and finished products, animals for breeding and fattening and other livestock; purchased stocks: goods and services purchased for resale, raw material and consumables).
(NSDCP) 2239 ‘Annual performance survey’ and (NSDCP) 1846 ‘Survey of local units’, (NSDCP) 1852 ‘Survey of local units of retail trade and catering services’: ‘Wages and salaries’ must be equal.
(NSDCP) 2239 ‘Annual performance survey’ and (NSDCP) 2241 ‘Annual labour survey’: ‘Regularly paid wages and salaries’ (payments of 1-12 months) must be equal, while the ‘accrued expenses of irregular payments’ (13th months salary, bonuses or other kind of lump-sum payments) could result in a difference as in case of labour statistics the actual time of the payment has to be taken into consideration while the performance statistics is accrual-based.
(NSDCP) 2239 ‘Annual performance survey’ and (NSDCP) 2240 ‘Annual investment survey’: ‘value of capitalized production’ must be consistent.
(NSDCP) 2239 ‘Annual performance survey’ and (NSDCP) 2146 ‘Sales structure of business services’: ‘value of turnover’ must be equal.
(NSDCP) 2239 ‘Annual performance survey’ and (NSDCP) 1039 ‘Annual survey of industrial products’: ‘value of turnover’, ‘export revenue’ and ‘excise duty’ must be equal.
(NSDCP) 2241 ‘Annual labour survey’ and (NSDCP) 1117 ‘Labour cost survey’: ‘value of wages and salaries’, ‘number of persons employed’ and ‘personnel cost’ must be consistent.
(NSDCP) 2240 ‘Annual investment survey’ and (NSDCP) 1933 ‘Annual survey of investment structure’: ‘Gross investment in tangible goods’ must be equal.
(NSDCP) 2414 check of comparison to the total data of the legal units through other HCSO Questionnaires [(NSDCP) 2239 'Annual performance survey', (NSDCP) 2240 'Annual investment survey'] and the taxation data we get twice every year from the National Tax and Customs Administration of Hungary (NTCA/NAV). Also check of data of partners (other legal units of the enterprise).
Surveys issued not for the primary purpose of observing of some indicator of income statement or balance sheet (turnover or revenues, production costs, personnel costs, investment) are not involved in the process of consistency checking of SBS data.
There are no more consistency checks other than the ones listed above.
15.4. Coherence - internal
Aggregated data are always consistent with their sub-aggregates as the aggregates are calculated from the micro level data.
Cost (total cost of a given data collection (EUR) /time expenditure (hour))
Performance, labour and investment statistics: 13.15
Regional statistics: 10.55
Environmental statistics: 13.35
Burden (number of questionnaires) * (average time for filling in the questionnaire (min))/60):
Performance, labour and investment statistics: 26416
Regional statistics: 2856
Environmental statistics: 6203
17.1. Data revision - policy
Annex I-IV:
There are no planned revisions.
Before the publication we correct the errors detected. After the publication we do not check the data, but if we notice a remarkable error during data processing of the following year, we correct the error of the previous year.
For reference year 2015 data revision for two regional series were carried out due to technical (IT) problems.
17.2. Data revision - practice
As there are no estimated part of the SBS reference population (every unit is included with its actual data in the preliminary dataset as well as in definitive data) no planned revision is applied.
Definitive data replace preliminary data as soon as they are available. Existing differences between these two datasets are due to the fact that HCSO receives the administrative dataset twice a year (1st: t+8 months, 2nd: t+12 months), and the second one contains updated information for some companies (e.g. companies implement self-correction and file a new tax return). This time the validation (survey data vs taxation data) is carried out repeatedly which can result in the correction of the survey data (with the cooperation of the affected unit).
The final data are not changed after publication.
18.1. Source data
Main input sources of SBS (detailed information see below):
1)
Annual business surveys (National Data Collection Programme nr 1845, 2239, 2240, 2241, 2414)
Regional surveys (National Data Collection Programme nr 1846, 1852)
Environmental surveys (National Data Collection Programme nr 1799, 1892)
2) Administrative data: taxation data and annual report data
Micro level data →Used for mass-imputation (legal units less than 20 number of persons employed and non-respondent legal units)
3) Additional data sources:
Legal units with 5-19 number of persons employed
No grossing up
Part of the investment and the labour statistics surveys
Primarily used for calculating proportions (parameters) (for estimation purposes)
Administrative data source covers the following types of tax-returns:
Tax return on non-flat rate tax of sole proprietors
Tax return on flat rate tax of sole proprietors
Tax return on simplified entrepreneurial tax of corporations with or without legal entity
Fixed-rate tax of low tax-bracket enterprises
Company income tax return of double-entry bookkeepers
Company income tax returns of double-entry bookkeepers choosing simplified entrepreneurial tax for next taxation year
Tax return on simplified entrepreneurial tax of corporations with or without legal entity
Small Business Tax of enterprises
The frequency of updating of administrative data sources: Good
Administrative data are subject to revision with (increasing) degree of completeness: Yes
Differences between the preliminary and the final datasets are due to the fact HCSO receives the administrative dataset twice a year (1st: t+8 months, 2 nd: t+12 months), and the second one contains more companies and updated information for some companies. For reference year 2020 the increase in the number of legal units were +17% (definitive data/preliminary data), and in the turnover: +10% in case of the administrative data. The reason for the larger discrepancy between final and preliminary data is that the tax data were sent later by data providers due to the covid epidemic.
Information on function of sample:
Sample used in size categories 0-4 and 5-9 are not functioning as real samples anymore (in the very beginning when the administrative data sources were unavailable samples were used). It means that there is no grossing up, the samples are used only for estimation purposes (in the lower size categories only administrative data sources can be used but not all of them is sufficiently detailed so estimations are done, e.g. in the case of ‘Tax return on non-flat rate tax of sole proprietors’ where the amount of material cost and the services come together and to define the above mentioned variables separately we have to rely on estimates).
Surveys (full scope and for sampling):
Labour, performance and investments variables:
1) (NSDCP) 2239 ‘Annual performances and expenses of enterprises’
NACE Rev. 2: organisation designated from section A-J, L-N, P-S, legal form 1, 2, 572, 573, 575, 576, 7 )
2) (NSDCP) 2240 ‘Annual investment survey’
Annual investment survey: all sections (NACE Rev. 2), legal form 1, 2, 572, 573, 575, 576, 7 with at least 20 persons employed (below 20 persons there is a sample).
3) (NSDCP) 2241 ‘Annual labour survey’
Annual labour survey: all sections (NACE Rev. 2), legal form 1, 2, 7 with at least 50 persons employed (there is a special sample for legal units with 5-49 persons employed).
4) (NSDCP) 2414 ‘Annual business survey of enterprises’
Enterprises with at least 4 legal units, with at least 500 million HUF aggregated net turnover from previous year) (TEÁOR’08/NACE Rev.2: organisation designated from section A-J, L-N, P-S, legal form 1, 2, 572, 573, 575, 576, 7)
Representative sample surveys without grossing up:
1) (NSDCP) 1845 ‘Annual business statistical survey’
Covers legal units with 0-4 number of persons employed
Contains labour, performances and investment characteristics
Type: stratified sample
Characteristics used for stratification:
Size class, activity (based on the frozen BR) and type of tax return (based on the information of the previous year):
Size class 0-4 in NACE Rev. 2 sections F, G, H, I, J in case legal units file ’Company income tax return of double-entry bookkeepers’.
Size class 0-4 in NACE Rev. 2 sections L, M, N in case of legal units file ’Company income tax return of double-entry bookkeepers’ or ’Tax return on simplified entrepreneurial tax of corporations with or without legal entity.
The effective sample size (%) for SBS population: 2.4%
As it is not functioning as a real sample, i.e. no grossing up is carried out, the calculation of effective sample size for each stratum is considered unnecessary.
2) Part of (NSDCP) 2240 ‘Annual investment survey’ and (NSDCP) 2241 ‘Annual labour survey’.
Cover legal units with 5-19 number of persons employed. From the reference year 2019 the NSDCP 2241 has a sample for the legal unit with 20-49 number of persons employed as well.
Type: stratified sample
Characteristics used for stratification:
Size class, activity (based on the frozen BR) and type of tax return (based on the information of the previous year):
Size class 5-19 in NACE Rev. 2 sections A-S in case of legal units file ’Company income tax return of double-entry bookkeepers’ or ‘Tax return on non-flat rate tax of sole proprietors’
The sample size (%) for SBS population: 11.1% for the legal units with 5-19 number of persons employed.
The sample size (%) for SBS population: 53.7% for the legal units with 20-49 number of persons employed.
As it is not functioning as a real sample, i.e. no grossing up is carried out, the calculation of effective sample size for each stratum is considered unnecessary.
Samples ( 1) and 2) ) are selected every 3 years and updated annually.
Regional variables: full scope surveys
Cover legal units with > 19 number of persons employed
1) (NSDCP) 1846: Survey of local units - covers NACE Rev.2: organisations designated from Sections B-E, H, J, K, L-N, P-S95-96, legal form 1, 2, 572, 573, 575, 576, 7 with at least 20 persons employed and Section F, legal form 1, 2, 572, 573, 575, 576, 7) with at least 10 persons employed.
2) (NSDCP) 1852: Survey of local units of retail trade and catering services - covers NACE Rev.2: organisations designated from Sections G, I, legal form 1, 2, 7 with at least 20 persons employed.
Environmental variables:
Representative sample survey (stratified sample):
1) one part of the (NSDCP) 1799 ‘Report on the environmental protection expenditures and on the environmental protection investments’
Characteristics used for stratification:
Size class, activity, region (Budapest and county):
size classes: 5-9, 10-19 in section D and in division E36
size classes: 5-9, 10-19, 20-49, 50-99 in sections: B, C
size classes: 20-49, 50-99 in sections: A, F-J, L-N, P-S
The effective sample size (%) for SBS population :
Total: 12.89% - For details see the annex (excel file) attached
Sample are selected and updated annually.
Full scope surveys:
2) Other part of NSDCP 1799 ‘Report on the environmental protection expenditures and on the environmental protection investments’ (not covered by the representative sample)
3) NSDCP 1892 ‘Report on the expenditures on environment protection, the production and the provision of environmental products and services’
One legal unit (reporting unit) is considered as one enterprise.
Frame on which the source is based: The Hungarian Business Register is the frame of SBS and all of the surveys used for SBS purposes.
For sampling and specifying the survey population a frozen status (concerning 31st of December) of register is used.
List of characteristics directly available or with a good proxy in the administrative source:
According to this methodology the following rule for stability is applied:
“Whenever a unit performs two activities both contributing to around 50% of the value added, a stability rule has been established in order to avoid frequent changes not reflecting a substantial change in the economic reality. According to this rule, the change of principal activity should be made when the current one has been accounting for less than 50% of the value added for at least two years.”
To define the principal activity, the top-down approach is applied.
Surveys: Data collection are carried out at several stages. Data collection is based on on-line surveys (electronic data collection system (ELEKTRA) of HCSO). The survey planning including the set of error checking rules is the task of the statisticians. The IT system sends an automatic e-mail to data providers if a new questionnaire becomes accessible, and if the deadline is expired.
Actions taken to speed up or increase the rate of response:
Re-contact by phone;
Sending a warning message by e-mail.
Only large businesses (legal units with more than 19 number of persons employed) are surveyed while data of small businesses are available form administrative data source (tax-return data).
Units having a fiscal year different from the calendar year and all the legal units applying IFRS are surveyed as well independently of size-class.
Administrative data:
Taxation data files are transferred by the National Tax and Customs Administration of Hungary via the integrated data transmission system of HCSO.
From the reference year 2020, the annual report data of the legal entities submitting simplified tax return are also used to impute their performance data. Annual report data are provided by the Ministry of Justice (via the integrated data transmission system of HCSO).
Data transfer of the administrative data is based upon a signed contract (data transfer agreement).
18.4. Data validation
Data are collected by electronic questionnaires. We receive all data via an electronic data collection system (ELEKTRA) and are transferred automatically into Integrated Data Entry and Validation System (ADÉL). The data are checked by the responsible statisticians in the ADÉL system. They can also give feedback for the data providers, and ask for corrections in case of errors.
Phases of data validation:
1st stage: Error checks and validations integrated into the data collecting system are run automatically during the filling of each questionnaire.
The data collection system (ELEKTRA) includes number of embedded automated error-checking and validation routines. However, only the significant errors are integrated into Electra which must be corrected by the data supplier (e.g. the rows „of which” cannot exceed the value of its total). If the data supplier does not correct the error, the questionnaire could not be finished and sent to HCSO.
2nd stage:
After the surveys are received by HCSO they are loaded into the data entry and validation system (ADÉL). This program carries out several pre-defined error-checks. When incorrect data are identified data suppliers are contacted individually and given the opportunity either to correct and re-send the data or confirm the original.
Most of the data-checks observes:
if a given value falls outside of the pre-defined range (minimum and maximum values),
if there is a significant difference (increase or decrease) between the reference year data and the previous year’s data,
if different consistencies are fulfilled within the survey (e.g. if there is a turnover data costs must be also provided),
if different consistencies are fulfilled among different surveys (these data-checks primarily observes data with similar methodological content in different surveys). Surveys used for cross-checking data: Infra-annual surveys (performance, labour, investment); Prodcom (Annual survey of industrial products), Survey of local units (regional data).
3rd stage: Comparison with administrative data (data of tax returns and financial statements)
When the administrative data will be available (8 months after the reference period) the survey data are checked with them. The administrative data provide a wide range of the variables available in the survey, especially in the case of doubly-entry bookkeepers (they present the about the 95-97% of the performance data in case of turnover, production value and value added).
In case of large companies the data of financial statements are also used for validation.
Cross-survey checking are carried out repeatedly in this period. The following surveys are involved in the validation: Prodcom (Annual survey of industrial products), Survey of local units (for regional data), Labour cost survey, Sales structure of business services). Data of foreign trade statistics (especially turnover, turnover related taxes and export revenues) are used for validation as well.
The output data (the SBS variables) are checked at individual and at aggregated level as well, both at legal unit and enterprise level.
For individual data checking the SQL Developer, the SAS and the own developed FORM-IT application are used.
Short description and applicability of FORM-IT app:
On the surface of the FORM of the Joint Database of legal units statisticians can see 16 significant variables which are available from the survey and from the tax return as well and have the same content. There are some thresholds referring to the maximum differences allowed between the two sources.
The colleagues observe the data of the two sources and if there is a difference between them, they can decide which one is the correct and chose the appropriate data source. (To decide which data is correct, the financial statements can be checked and/or a contact can be made to the given legal unit. If the data is incorrect, the data provider has to re-send the questionnaire.)
Differences more than 500 million HUF have to be corrected in the Data entry validation system and can’t be performed by simply changing the source.
The colleagues can checkmark the verified data.
The aggregated level data are checked with a general tabulation program – within the JPD system (SBS database and system).
Statisticians can query tables according to predefined dimension e.g. legal form, size category etc. or relations between the data (e.g. capitalised production – investment) and can compile comparative table for the preliminary and definitive data.
Taxation data are corrected by the NA department in the taxation databases.
18.5. Data compilation
The data of non-responding units (0.82%) at legal unit level are imputed at micro level on the basis of previous years’ data. These data are not adjusted with ratio of the year/year average change.
No grossing up is used as we have micro-data for the total population.
18.6. Adjustment
Units having a fiscal year different from the calendar year are surveyed as well independently of size-class. This way we receive data for the calendar year begins 1st of January and ends 31st of December.
No other comments.
Structural business statistics (SBS) describes the structure, conduct and performance of economic activities, down to the most detailed activity level (several hundred economic sectors).
SBS covers all activities of the non-financial business economy with the exception of agricultural activities and personal services. Limited information is available on banking, insurance and pension funds.
Main characteristics (variables) of the SBS data category:
Business demographic variables (e.g. number of enterprises),
"Output related" variables (e.g. turnover, value added),
"Input related" variables: labour input (e.g. employment, hours worked); goods and services input (e.g. total of purchases); capital input (e.g. material investments).
Reporting and observation unit: legal and enterprise unit.
From the reference year 2018 the scope of statistics broadened with a new observation unit: the statistical unit enterprise. Monitoring and consolidating the internal flow between individual members of the enterprise makes it possible to measure the process of the national value creation more accurately.
There are three groups of enterprises:
SIMPLE ENTERPRISES: The legal entity is not related to another company, i.e. an enterprise unit can be matched to a legal entity.
COMPLEX SMALL ENTERPRISES: for the enterprises created by linking 2-3 legal units, an automatic consolidation algorithm has been developed in accordance with international recommendations.
COMPLEX LARGE ENTERPRISES: for enterprises with a minimum of 4 legal units, a questionnaire is used to observe each member of the enterprise and to compile data for internal flow. For this purpose the questionnaire 2414 (‘Survey on Enterprises’ - enterprises with at least 4 legal units, where the aggregate turnover of the legal units in the previous period reaches HUF 500 million) has been developed.
SBS population is identified on the base of the Business Register. The Business Register provides the frame of the SBS: the legal units /enterprise statistical units for which statistical survey (performance, labour, investment) or administrative data are available, will be involved in SBS population. None of the NACE activity codes or size classes are excluded from the target population.
SBS data covers the territory of Hungary. All regions are covered by the SBS data.
SBS reference population includes foreign branches of legal units registered in the territory of Hungary which do not constitute a separate legal entity and are dependent on the resident legal unit.
Branches of foreign legal units are involved in SBS as well, while the performance (output) related activities abroad of Hungarian legal units (foreign branches of resident units which constitute a separate legal entity) are excluded according to the Regulation (EC) No 295/2008.
Accordingly, Hungarian SBS data is consistent with ESA2010 and BPM6 international regulations requirements.
2020
SBS data refers to calendar year. Units are known applying a fiscal year different from the calendar year are surveyed (regardless of the size-class).
Sampling, coverage and measurement errors: not applicable
Only the non-response rate (0.82 percent in case of final SBS data at legal unit level) has a limited effect on the overall accuracy.
Preliminary data are published on 15th of November (in year t+1), while the final data are released in the end of August (in year t+2).
Number of enterprises and number of local units are expressed in units.
Monetary data are expressed in millions of €.
Employment variables are expressed in units.
Per head values are expressed in thousands of € per head.
Ratios are expressed in percentages.
The data of non-responding units (0.82%) at legal unit level are imputed at micro level on the basis of previous years’ data. These data are not adjusted with ratio of the year/year average change.
No grossing up is used as we have micro-data for the total population.
Main input sources of SBS (detailed information see below):
1)
Annual business surveys (National Data Collection Programme nr 1845, 2239, 2240, 2241, 2414)
Regional surveys (National Data Collection Programme nr 1846, 1852)
Environmental surveys (National Data Collection Programme nr 1799, 1892)
2) Administrative data: taxation data and annual report data
Micro level data →Used for mass-imputation (legal units less than 20 number of persons employed and non-respondent legal units)
3) Additional data sources:
Legal units with 5-19 number of persons employed
No grossing up
Part of the investment and the labour statistics surveys
Primarily used for calculating proportions (parameters) (for estimation purposes)
Administrative data source covers the following types of tax-returns:
Tax return on non-flat rate tax of sole proprietors
Tax return on flat rate tax of sole proprietors
Tax return on simplified entrepreneurial tax of corporations with or without legal entity
Fixed-rate tax of low tax-bracket enterprises
Company income tax return of double-entry bookkeepers
Company income tax returns of double-entry bookkeepers choosing simplified entrepreneurial tax for next taxation year
Tax return on simplified entrepreneurial tax of corporations with or without legal entity
Small Business Tax of enterprises
The frequency of updating of administrative data sources: Good
Administrative data are subject to revision with (increasing) degree of completeness: Yes
Differences between the preliminary and the final datasets are due to the fact HCSO receives the administrative dataset twice a year (1st: t+8 months, 2 nd: t+12 months), and the second one contains more companies and updated information for some companies. For reference year 2020 the increase in the number of legal units were +17% (definitive data/preliminary data), and in the turnover: +10% in case of the administrative data. The reason for the larger discrepancy between final and preliminary data is that the tax data were sent later by data providers due to the covid epidemic.
Information on function of sample:
Sample used in size categories 0-4 and 5-9 are not functioning as real samples anymore (in the very beginning when the administrative data sources were unavailable samples were used). It means that there is no grossing up, the samples are used only for estimation purposes (in the lower size categories only administrative data sources can be used but not all of them is sufficiently detailed so estimations are done, e.g. in the case of ‘Tax return on non-flat rate tax of sole proprietors’ where the amount of material cost and the services come together and to define the above mentioned variables separately we have to rely on estimates).
Surveys (full scope and for sampling):
Labour, performance and investments variables:
1) (NSDCP) 2239 ‘Annual performances and expenses of enterprises’
NACE Rev. 2: organisation designated from section A-J, L-N, P-S, legal form 1, 2, 572, 573, 575, 576, 7 )
2) (NSDCP) 2240 ‘Annual investment survey’
Annual investment survey: all sections (NACE Rev. 2), legal form 1, 2, 572, 573, 575, 576, 7 with at least 20 persons employed (below 20 persons there is a sample).
3) (NSDCP) 2241 ‘Annual labour survey’
Annual labour survey: all sections (NACE Rev. 2), legal form 1, 2, 7 with at least 50 persons employed (there is a special sample for legal units with 5-49 persons employed).
4) (NSDCP) 2414 ‘Annual business survey of enterprises’
Enterprises with at least 4 legal units, with at least 500 million HUF aggregated net turnover from previous year) (TEÁOR’08/NACE Rev.2: organisation designated from section A-J, L-N, P-S, legal form 1, 2, 572, 573, 575, 576, 7)
Representative sample surveys without grossing up:
1) (NSDCP) 1845 ‘Annual business statistical survey’
Covers legal units with 0-4 number of persons employed
Contains labour, performances and investment characteristics
Type: stratified sample
Characteristics used for stratification:
Size class, activity (based on the frozen BR) and type of tax return (based on the information of the previous year):
Size class 0-4 in NACE Rev. 2 sections F, G, H, I, J in case legal units file ’Company income tax return of double-entry bookkeepers’.
Size class 0-4 in NACE Rev. 2 sections L, M, N in case of legal units file ’Company income tax return of double-entry bookkeepers’ or ’Tax return on simplified entrepreneurial tax of corporations with or without legal entity.
The effective sample size (%) for SBS population: 2.4%
As it is not functioning as a real sample, i.e. no grossing up is carried out, the calculation of effective sample size for each stratum is considered unnecessary.
2) Part of (NSDCP) 2240 ‘Annual investment survey’ and (NSDCP) 2241 ‘Annual labour survey’.
Cover legal units with 5-19 number of persons employed. From the reference year 2019 the NSDCP 2241 has a sample for the legal unit with 20-49 number of persons employed as well.
Type: stratified sample
Characteristics used for stratification:
Size class, activity (based on the frozen BR) and type of tax return (based on the information of the previous year):
Size class 5-19 in NACE Rev. 2 sections A-S in case of legal units file ’Company income tax return of double-entry bookkeepers’ or ‘Tax return on non-flat rate tax of sole proprietors’
The sample size (%) for SBS population: 11.1% for the legal units with 5-19 number of persons employed.
The sample size (%) for SBS population: 53.7% for the legal units with 20-49 number of persons employed.
As it is not functioning as a real sample, i.e. no grossing up is carried out, the calculation of effective sample size for each stratum is considered unnecessary.
Samples ( 1) and 2) ) are selected every 3 years and updated annually.
Regional variables: full scope surveys
Cover legal units with > 19 number of persons employed
1) (NSDCP) 1846: Survey of local units - covers NACE Rev.2: organisations designated from Sections B-E, H, J, K, L-N, P-S95-96, legal form 1, 2, 572, 573, 575, 576, 7 with at least 20 persons employed and Section F, legal form 1, 2, 572, 573, 575, 576, 7) with at least 10 persons employed.
2) (NSDCP) 1852: Survey of local units of retail trade and catering services - covers NACE Rev.2: organisations designated from Sections G, I, legal form 1, 2, 7 with at least 20 persons employed.
Environmental variables:
Representative sample survey (stratified sample):
1) one part of the (NSDCP) 1799 ‘Report on the environmental protection expenditures and on the environmental protection investments’
Characteristics used for stratification:
Size class, activity, region (Budapest and county):
size classes: 5-9, 10-19 in section D and in division E36
size classes: 5-9, 10-19, 20-49, 50-99 in sections: B, C
size classes: 20-49, 50-99 in sections: A, F-J, L-N, P-S
The effective sample size (%) for SBS population :
Total: 12.89% - For details see the annex (excel file) attached
Sample are selected and updated annually.
Full scope surveys:
2) Other part of NSDCP 1799 ‘Report on the environmental protection expenditures and on the environmental protection investments’ (not covered by the representative sample)
3) NSDCP 1892 ‘Report on the expenditures on environment protection, the production and the provision of environmental products and services’
One legal unit (reporting unit) is considered as one enterprise.
Frame on which the source is based: The Hungarian Business Register is the frame of SBS and all of the surveys used for SBS purposes.
For sampling and specifying the survey population a frozen status (concerning 31st of December) of register is used.
List of characteristics directly available or with a good proxy in the administrative source:
According to this methodology the following rule for stability is applied:
“Whenever a unit performs two activities both contributing to around 50% of the value added, a stability rule has been established in order to avoid frequent changes not reflecting a substantial change in the economic reality. According to this rule, the change of principal activity should be made when the current one has been accounting for less than 50% of the value added for at least two years.”
To define the principal activity, the top-down approach is applied.
Dissemination database tables (both at legal unit and enterprise level) are published 10,5 months after the reference period (in case of preliminary data), while the second publishing (definitive data) is by the end of August (which means 20 months after the reference year).
The whole country is treated in a uniform manner from a statistical point of view. The same statistical concepts are applied for the entire territory of Hungary. There are no difficulties concerning the geographical comparability.
2008-2017
2018-2020
In 2008 the Nace Rev. 2 has been introduced, so the time-series broken down by activity are not comparable with the former data series.
There is a break in the series in 2018 due to the change of the reporting and observation unit. Until 2018, the HCSO reported data at the level of legal units.
