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.
Eurostat, the statistical office of the European Union
1.2. Contact organisation unit
G2: European businesses
1.3. Contact name
Restricted from publication
1.4. Contact person function
Restricted from publication
1.5. Contact mail address
2920 Luxembourg LUXEMBOURG
1.6. Contact email address
Restricted from publication
1.7. Contact phone number
Restricted from publication
1.8. Contact fax number
Restricted from publication
2.1. Metadata last certified
28 January 2020
2.2. Metadata last posted
28 January 2020
2.3. Metadata last update
28 January 2020
3.1. Data description
The Community Innovation Survey (CIS) is a survey of innovation activity in enterprises. The harmonised survey is designed to provide information on the innovativeness of sectors by type of enterprises, on the different types of innovation and on various aspects of the development of an innovation, such as the objectives, the sources of information, the public funding or the expenditures.
The CIS provides statistics broke down by countries, type of innovators, economic activities and size classes. The survey is currently carried out every two years across the EU Member States, some EFTA countries and EU candidate countries.
In order to ensure comparability across countries, Eurostat, in close cooperation with the countries, has developed a standard core questionnaire starting with the CIS3 data collection, along with an accompanying set of definitions and methodological recommendations. The concepts and underlying methodology of the CIS are also based on the Oslo Manual — second edition of 1997 and third edition of 2005 (see link at the bottom of the page).
The data presented in these tables refer to enterprises with ‘10 employees or more’ active in the sectors to be covered under the Regulation (cf. NACE CORE). Further activities may be covered on a voluntary basis. Most statistics are based on a reference period of three years, but some use one calendar year.
Since CIS 2008, the survey has included an ad-hoc module. It consists of a set of questions focusing on a special theme. The themes are different in each survey wave, allowing data to be obtained on specific issues beyond the data usually collected.
Overview over time:
Initially, the CIS data collection was carried out every four years. The first collection (CIS Light) was launched in 1993 as a pilot exercise and the second (CIS2) was carried out in 1997/1998 for most countries except Greece and Ireland, where it was launched in 1999. The third survey (CIS3) was conducted in 2000/2001 for most participating countries with the exception of Norway, Iceland, Luxembourg and Greece, where it was launched in 2002.
As from 2004, the survey has been carried out every two years.
CIS4 was conducted in the EU Member States (as for 2004), Iceland, Norway, Bulgaria and Romania. The survey was launched in 2005 with a three-year reference period 2002 to 2004 for most indicators.
The fifth survey CIS 2006 was carried out in all EU Member States (as for 2006), Norway, Bulgaria, Romania, Croatia and Turkey. It was launched in 2007, mostly for the reference period 2004 to 2006.
As regards CIS 2008, EU Member States (all except Greece), Iceland, Norway, Croatia and Turkey took part in the survey. CIS 2008 was launched in 2009 with a three-year reference period 2006 to 2008 for most indicators. Changes were made to the CIS 2008 questionnaire to bring it into line with the third revision of the Oslo Manual, 2005 edition, by giving greater weight to organisational and marketing innovation. CIS 2008 was complemented by an ad-hoc module on innovation with environmental benefits.
The seventh Community Innovation Survey, CIS 2010, had 31 participating countries (all EU Member States except Greece, Iceland, Norway, Croatia, Serbia and Turkey) and reported most results for the reference period 2008-2010. CIS 2010 also follows the recommendations of the Oslo Manual and reports indicators on four types of innovation: product, process, organisational and marketing.
However, despite implementation of the recommendations of the third edition of the Oslo Manual, the question on innovation expenditures is still limited to product and process innovation in order to maintain continuity with earlier versions of the CIS. Furthermore, generally fewer questions are asked about organisational and marketing innovation than about product and process innovation.
While the European innovation statistics use the aggregated national data, the microdata sets can be accessed by researchers via the SAFE Centre of Eurostat in Luxembourg or via the microdata on CD-ROM releases in more anonymised form; some countries also provide access to their micro-data at similar safe centres.
3.2. Classification system
Enterprises are classified by country, economic activity (NACE), size class and type of innovation activity.
3.3. Coverage - sector
The changes to the Statistical Classification of Economic Activities in the European Community entail changes to categories in the Community Innovation Survey as well as decisions on mandatory (core) coverage.
From CIS 2 to CIS2006, the activity classification was NACE Rev 1.1, the classification in force at the time. The following industries under this classification were — in general — included in the CIS population:
Core coverage
mining and quarrying (NACE C10-C14)
manufacturing (NACE D15-D37)
electricity, gas and water supply (NACE E40-E41)
wholesale trade and commission trade, except of motor vehicles and motorcycles (NACE G51)
transport, storage and communication (NACE I60-I64)
financial intermediation (NACE J65-J67)
computer and related activities (NACE K72)
architectural and engineering activities (NACE K74.2)
technical testing and analysis (NACE K74.3)
Additional coverage was provided on a voluntary basis for CIS 4 and CIS 2006, including the following sectors: research and development (NACE K73)/ construction (NACE F45)/ sale, maintenance and repair of motor vehicles and motorcycles; retail sale of automotive fuel (NACE G50)/ retail trade, except of motor vehicles and motorcycles; repair of personal and household goods (NACE G52)/ hotels and restaurants (NACE G55)/ real estate activities (NACE K70)/ renting of machinery and equipment without an operator (NACE K71)/ all other business services (NACE K74.1 to K74.8).
In contrast to previous CIS surveys that used NACE Rev. 1.1, CIS 2008 and CIS 2010 used the NACE Rev. 2 classification of economic activities.
Core coverage
As required by Annex IV to Commission Regulation No 973/2007, the following NACE Rev.2 divisions were included in the core target population of CIS 2008 and CIS 2010:
mining and quarrying (NACE B05-B09)
manufacturing (NACE C10-C33)
electricity, gas steam and air conditioning supply (NACE D35)
water supply; sewerage, waste management and remediation activities (NACE E36-E39)
wholesale trade, except of motor vehicles and motorcycles (NACE G46)
transportation and storage (NACE H49-H53)
publishing activities (NACE J58); telecommunications (NACE J61); computer programming, consultancy and related activities (NACE J62)
information services activities (NACE J63)
financial and insurance activities (NACE K64-K66)
architectural and engineering activities; technical testing and analysis (NACE M71)
Additional coverage on a voluntary basis includes the following sectors: agriculture, forestry and fishing (NACE A01-A03)/ scientific research and development (NACE M72)/ construction (NACE F)/ wholesale and retail trade and repair of motor vehicles and motorcycles (NACE G45)/ retail trade, except of motor vehicles and motorcycles (NACE G47)/ accommodation and food service activities (NACE I55-I56)/ motion picture, video and television programme production, sound recording and music publishing-programming and broadcasting activities (NACE J59-J60)/ real estate activities (NACE L68)/ legal and accounting activities (NACE M69)/ activities of head offices; management consultancy activities (NACE M70)/ advertising and marketing research (NACE K73)/ other professional, scientific and technical activities (NACE K74)/ veterinary activities (NACE M75)/ rental and leasing activities (NACE N77)/ employment activities (NACE N78)/ travel agency, tour operator and other reservation services and related activities (NACE N79)/ security and investigation activities (NACE N80)/ services to buildings and landscape activities (NACE N81)/ office administrative, office support and other business support activities (NACE N82)
3.4. Statistical concepts and definitions
CIS provides information on the characteristics of innovation activity at enterprise level. It allows Europe’s progress to be monitored in the area of innovation, creating a better understanding of the innovation process with analysis of the objectives and the effects of innovation. These results can also be linked to variables related to competitiveness, employment and economic growth. The concepts are in line with those recommended by the Oslo Manual (2nd edition 1997 and 3rd edition 2005), which is the internationally recognised standard methodology for collecting innovation statistics.
From CIS3 to CIS2006, the innovation concept was still limited to technological innovation (product and process innovation). The following categories of indicators were provided for these three CIS waves.
Product, process, on-going and abandoned innovation
Innovation activity and expenditure
Intramural research and experimental development (R&D)
Effects of innovation
Public funding of innovation
Innovation cooperation
Sources of information for innovation
Hampered innovation activity
Patents and other protection methods
Other important strategic and organisational changes in the enterprise
From CIS 2008 onwards, information has been collected not only on product and process innovation but also on organisational and marketing innovation. Most questions cover new or significantly improved goods or services or the implementation of new or significantly improved processes, logistics or distribution methods. CIS results offer a broad set of indicators on innovation activities, innovation expenditures, public funding, sources of information for innovation, innovation cooperation, innovation objectives and organisational and marketing innovation.
Main concepts and definitions used for the CIS data collection:
An innovation is the implementation of a new or significantly improved product (good or service), process, new marketing method, or new organisational method in business practices, workplace organisation or external relations. This broad definition of innovation encompasses a wide range of possible innovations. An innovation can be more narrowly categorised as the implementation of one or more types of innovations, for instance product and process innovations. An overview of the types of innovators is available in annex.
The minimum requirement for an innovation is that the product, process, marketing method or organisational method must be new to the firm (or significantly improved). This includes products, processes and methods that firms are the first to develop and those that have been adopted from other firms or organisations.
Product innovative enterprises are those who introduced, during the period under review, new and significantly improved goods and/or services with respect to their fundamental characteristics, technical specifications, incorporated software or other immaterial components, intended uses, or user friendliness. Changes of a solely aesthetic nature and the simple resale of new goods and services purchased from other enterprises are not considered as innovation.
Process innovative enterprises implemented new and significantly improved production technologies or new and significantly improved methods of supplying services and delivering products during the period under review. The outcome of such innovations should be significant with respect to the level of output, quality of products (goods or services) or costs of production and distribution. Purely organisational or managerial changes are not included.
Organisational innovative enterprises implemented a new organisational method in the enterprise’s business practices, workplace organisation or external relations.
Marketing innovative enterprises implemented a new marketing method involving significant changes in product design or packaging, product placement, product promotion or pricing.
Innovative enterprises had innovation activities during the period under review, including enterprises with on-going and abandoned activities. In other words, enterprises that had innovation activities during the period under review, regardless of whether the activity resulted in the implementation of an innovation, are innovation-active.
During a given period, innovation activities can be of three kinds:
successful, in having resulted in the implementation of an innovation (although the innovation need not have been commercially successful);
on-going, with work in progress that has not yet resulted in the implementation of an innovation.
abandoned before the implementation of an innovation.
Non-innovative enterprises had no innovation activity whatsoever during the reference period. These enterprises answered only a limited set of questions from the survey in relation to the absence of innovation activity, factors hampering innovation, patents and other protection methods, etc. The CIS 2008 and CIS 2010 modules (on eco-innovation and on creativity and skills, respectively) targeted enterprises both with and without innovation activity.
3.5. Statistical unit
The statistical unit is the enterprise, as defined in the Council Regulation on statistical units (Council Regulation (EEC) No 696/93 of 15 March 1993, OJ No L76 of 3 March 1993) or as defined in the statistical business register.
The Regulation defines the enterprise as ‘the smallest combination of legal units that is an organisational unit producing goods or services, which benefits from a certain degree of autonomy in decision making, especially for the allocation of its current resources. It may carry out one or more activities at one or more locations and it may be a combination of legal units, one legal unit or part of a legal unit ’.
3.6. Statistical population
The population of the CIS is determined by the size of the enterprise and its principal activity. All enterprises with 10 or more employees in any of the specified sectors are included in the statistical population. For the coverage of economic activities, see section 3.3 (Sector coverage) above.
3.7. Reference area
Most EU Member States took part in CIS Light, CIS 2, CIS 3 and CIS 2004. The number of countries participating has increased over time. Additionally, some EFTA, EU candidate and non-member countries have conducted the survey as well.
CIS2006:
The fifth survey CIS 2006 was carried out in all EU Member States (as for 2006), Norway, Bulgaria, Romania, Croatia and Turkey.
CIS 2008:
The survey was carried out in all EU Member States except Greece. Iceland, Norway, Croatia and Turkey also took part.
CIS 2010
Thirty-one countries participated in the 2010 Community Innovation Survey: all EU Member States (except Greece), Iceland, Norway, Croatia, Serbia and Turkey.
3.8. Coverage - Time
The results of each Community Innovation Survey are available in separate sections in Eurobase. This is because the same indicators would not be comparable from a survey to another because of changes in the questionnaire.
3.9. Base period
Not applicable.
CIS data are available in the following units: Absolute value (NBR) and Percentage (PC).
All the financial variables (innovation expenditures and turnover) are presented in thousand euros.
A particular feature of the Community Innovation Survey is that it uses different reference periods (except CIS 2):
most indicators are based on a three-year reference period, for example, the indicator showing whether an enterprise introduced an innovation covers a period of three years;
a minority of indicators are based on one year. It's the case of the turnover and employment indicators, which are usually requested for the first and last years of the 3-year reference period. Innovation expenditure is also based only on the last year of this 3-year period.
CIS Light:
Countries were free to choose between two reference periods of 3 years: 2000-2002 or 2001-2003. Latvia, Slovakia and Norway had data for the period 2001-2003, while most other countries had data for the first period, 2000-2002, with the exception of Luxembourg, Lithuania, Malta and Hungary, which used a 2-year reference period (2002-2003), Slovenia, which used a 2-year reference period (2001-2002), and Portugal, which reported results for only a single reference year (2003).
CIS2
All data referred to the calendar year 1996.
CIS 3
The standard reference period for the indicators with a 3-year period was from 1998 to 2000.
Exceptions to the reference year in CIS3:
Poland generally had a reference period of 1998-2000 for the industrial sector and 1997-1999 for the services sector. The Czech Republic, Hungary, Latvia, Lithuania, Slovakia and Norway had a reference period of 1999-2001. Romania had a reference period of 2000-2002. Slovenia had a two-year reference period of 2001-2002, while Bulgaria had a reference period of 2001-2003. Spain used an earlier version of the CIS3 core questionnaire.
Moreover, there are some exceptions: the data on turnover (tables inn_bas and inn_prod) refer to 2001; the data on expenditure (table inn_exp) are for 2000 for the industry sector and for 1999 for the services sector; the data on non-patent protection methods (table inn_pat) refer to 1996; and the data on non-technological innovation (table inn_cha) refer to 2001.
CIS4
The standard reference period for the indicators with a 3-year period was from 2002 to 2004. Indicators such as innovation expenditures were based on the reference year 2004. The turnover and employment figures were reported for the calendar year 2002 and for the calendar year 2004. All countries collected the data in accordance with these reference periods, with just the Czech Republic having a reference period of 2003-2005.
CIS2006
The standard reference period for the indicators with a 3-year period was from 2004 to 2006. Indicators such as innovation expenditures were based on the reference year 2006, while the turnover and employment figures were reported for the calendar year 2004 and for the calendar year 2006.
CIS 2008
Similar to the previous surveys, three reference periods were used in the questionnaire: the reference period for most questions was 2006 to 2008, i.e. the three-year period from the beginning of 2006 to the end of 2008. The indicators on innovation expenditures were based on the calendar year 2008. Additionally, the turnover and employment of enterprises were requested for two years: 2006 and 2008.
All countries collected the data in accordance with these reference periods.
CIS 2010
Most CIS 2010 questions covered the reference period 2008 to 2010, i.e. the 3-year period from the beginning of 2008 to the end of 2010. As regards the indicators on innovation expenditures for product and process innovations, the reference period was only one year, 2010. Spain also used just 2010 as the reference period to collect the number of enterprises engaged in innovation activities. The other countries used the 3-year period for these indicators. Moreover, a limited number of basic economic indicators such as turnover and employment were requested for 2008 and for 2010.
6.1. Institutional Mandate - legal acts and other agreements
Both the CIS2 and CIS3 surveys were carried out under a gentleman’s agreement. However, a harmonised survey methodology was provided to the participating countries.
Microdata on CIS 3 are held at Eurostat to allow further data compilation and access in accordance with Commission Regulation 831/2002.
The CIS4 survey was based on Commission Regulation No 1450/2004, implementing Decision No 1608/2003/EC of the European Parliament and of the Council concerning the production and development of Community statistics on innovation (= Commission Regulation on innovation statistics). This Regulation puts innovation statistics on a statutory basis and makes the delivery of certain variables compulsory. In addition, it defines the obligatory sector coverage. It also applies to CIS 2006, CIS 2008 and CIS 2010.
6.2. Institutional Mandate - data sharing
Not applicable.
7.1. Confidentiality - policy
Regulation (EC) No 223/2009 on European statistics (recital 24 and Article 20(4)) of 11 March 2009 (OJ L 87, p. 164) refers to the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and access to those confidential data with due account for technical developments and the requirements of users in a democratic society.
7.2. Confidentiality - data treatment
Primary and secondary confidentiality of CIS data are flagged by the countries.
The objective is to protect cells against absolute and approximate disclosure but to also minimize information loss in order to increase the usefulness of these data. Eurostat takes the complete flagging as given and is not executing any additional measure than removing flagged data while publishing the results. Primary and secondary confidentiality is respected for any data publicly released.
The European aggregates are not made available if at least one country sent confiential data for this indicator.
8.1. Release calendar
The release calendar is publicly available and published on Eurostat’s website.
Since 2004, Community Innovation Surveys have been conducted every two years.
10.1. Dissemination format - News release
News releases on-line
10.2. Dissemination format - Publications
Pocketbook and Statistics Explained
10.3. Dissemination format - online database
Please consult free data on-line Eurobase / science and technology.
10.4. Dissemination format - microdata access
The CIS microdata sets are available in two versions differing on the level of information retained:
Version one: confidential data as obtained from the national authorities. They allow only indirect identification of the statistical units concerned. The access to these data is only possible in Eurostat SAFE Centre.
Version two: anonymised microdata sets extracted from the aforementioned data. The individual statistical records which have been modified in order to minimise, in accordance with current best practice, the risk of identification of the statistical units to which they relate. This access is done via distribution of encrypted CD-ROM according to contracts established between Eurostat and the corresponding research institutions.
The implementation frameworks for release of EU microdata for scientific purposes are defined in the Commission Regulation (EC) No 831/2002.
As regards methodology, Commission Regulation No 1450/2004 provides details of the compilation of innovation statistics. In accordance with the Oslo Manual, it provides the harmonised concepts and definitions that Member States should apply for compiling the statistics.
10.7. Quality management - documentation
Detailed information on the quality of CIS 2006 and CIS 2008 can be found in the ‘Synthesis Quality Report’ (See annex at the bottom of the page).
11.1. Quality assurance
In accordance with section 7(4) of the Annex to Commission Regulation No 1450/2004 on innovation statistics, metadata should also be sent as part of the Quality Report. This will include key quality indicators such as non-response rates, coefficient of variation, etc.
The CIS Quality Report template has been developed in agreement with the participating countries. It conforms to the ESS Standard for Quality Reports as well as the methodological concepts and requirements set out in the Oslo Manual and the EU legislation on innovation statistics.
11.2. Quality management - assessment
A high degree of harmonisation in concepts and methods has been achieved through the adoption of Commission Regulation No 1450/2004 and the Oslo Manual for the compilation of innovation statistics. However, countries may deviate from the harmonised CIS questionnaire by introducing additional questions and/or modifying the existing questions in order to satisfy specific national needs.
12.1. Relevance - User Needs
The CIS provides an understanding of the innovation process, sources of information, cooperation, and objectives of enterprises, and allows some cross-analyses using some economic indicators such as the number of employees or turnover. Analyses are supported by various EU RTD Framework Programmes. Since 2000, the CIS has also become a major data source for the ‘European Innovation Scoreboard’ and the ‘Regional Innovation Scoreboard’.
12.2. Relevance - User Satisfaction
A user satisfaction survey is not generally carried out and countries mainly receive ad-hoc feedback from users.
12.3. Completeness
The completeness of national statistics is assessed by a comparison between the number of compulsory cells to be transmitted to Eurostat and the number of cells actually transmitted. All countries transmitted all the compulsory statistics on innovation in the 2010 survey.
13.1. Accuracy - overall
The overall picture is that all countries make considerable efforts to reduce errors or at least to identify and correct them. Interviewer training and assistance to respondents are provided for data collection. Comprehensive data validation is the norm during and after data collection. Re-contacting enterprises is the primary option for correcting errors, so the use of imputation is less needed. All these factors point to the high quality of the innovation statistics.
13.2. Sampling error
Appropriate stratification yields results with smaller sampling errors and makes it possible to ensure that there are enough units in the respective domains to produce results of good quality. However, according to the CIS 2008 Synthesis Quality Report, the coefficients of variation can be quite large in a few countries and for specific variables, and it seems that this is not always due to small sample rates.
13.3. Non-sampling error
Coverage errors are minimal in CIS 2008. Business registers are generally used as the sampling frame, offering up-to-date and high-quality data. Most errors reported arise from over-coverage due to the inclusion of out-of-scope units or units that no longer exist in the frame population.
Measurement errors are also very limited thanks to the continuous efforts of all countries to reduce such errors through testing of the questionnaire, training of the staff involved in the survey and the provision of detailed guidelines for completion of the CIS questionnaire.
As regards non-response errors, they may be reduced with the use of reminders/recalls to enterprises. As recommended in the survey guidelines, most countries send at least two or three paper reminders to non-responding enterprises. Additionally, these enterprises are contacted by phone or e-mail in order to remind them to fill in and deliver the survey questionnaire. Overall, response rates exceed 60 % with the exception of a few countries.
14.1. Timeliness
According to Commission Regulation No 1450/2004, national CIS statistics must be delivered to Eurostat within 18 months from the end of the reference year.
Most countries conform to the Regulation and delivered the data by the requested time.
The CIS2 data were released in the first quarter of 1999. The CIS3 data were disseminated by Eurostat in August 2003. The CIS4 data were disseminated in October 2006. The CIS 2006 data collection was disseminated in October 2008. The CIS 2008 data collection was released in November 2010. Eurostat made available CIS 2010 data at the beginning of November 2012.
14.2. Punctuality
The vast majority of countries are very punctual in delivering CIS data to Eurostat.
15.1. Comparability - geographical
In order to ensure comparability across countries, Eurostat, in close cooperation with the EU Member States and other countries, has developed a standard core questionnaire for each CIS round, with an accompanying set of definitions and methodological recommendations.
15.2. Comparability - over time
Comparability of CIS3 and CIS2 data
Comparability of data between the second and third Community Innovation Surveys is limited due to differences in survey methodology, target population, survey questionnaires and the definition of innovation. These differences concern:
Lower cut-off for inclusion in the target population
The cut-off for inclusion in the target population in CIS3 is firms with 10 employees and more in both the manufacturing and service sectors. In CIS2 the cut-off for inclusion in the target population is firms with 20 employees in the manufacturing sector and firms with 10 employees in the service sector.
More industries included in the target population
The CIS3 target population includes more activities defined by NACE than the CIS 2 survey. The additional activities covered by CIS3 are mining and quarrying (NACE 10-14), research and development (NACE 73) and technical testing and analysis (NACE 74.3). In addition, all NACE activities 63-64 (storage and communication) and the entire NACE activity 74.2 (architectural and engineering activities) were included.
Changed definition of innovation
The definition of innovation changed slightly in CIS3 compared to CIS2. In CIS3 the term ‘Innovation’ is used instead of ‘Technological innovation’. The term ‘technological’ remains though in the explanatory text of the concept.
Same core questionnaire used for manufacturing and services enterprises
In CIS3 a common core questionnaire for both kinds of enterprise (manufacturing and services) was used. This was not the case with CIS2, where two different core questionnaires were used for these two sectors. Enterprises in the service sector were asked approximately 90 per cent of the questions of the core questionnaire for enterprises in manufacturing. In the core questionnaire for service enterprises in CIS2 there was no split between product and process innovations, and no questions were asked on turnover from products new to the enterprise or turnover from products new to the market. The use of one common core questionnaire for CIS3 provides more harmonised data on overall innovation activities. The data are also more comparable between the two main sectors.
More information collected on both innovators and non-innovators
The CIS3 core questionnaire is substantially larger than the two core questionnaires used in CIS2. It therefore provides a more complete picture of both innovating and non-innovating enterprises. Under the assumption that the reporting enterprise was both product- and process-innovative, the CIS3 core questionnaire is approximately 50 per cent bigger than the CIS2 questionnaire in terms of the number of questions. More information is also collected on non-innovative enterprises in the CIS3 core questionnaire since non-innovative firms are asked more questions in CIS3 than in CIS2.
Both CIS Light and CIS2 were implemented on a voluntary basis by countries. Therefore no standardised methodology or questionnaire was implemented at national level. The differences in data collection may affect both inter-country and intra-country comparison, i.e. data may not be comparable either between countries or between CIS and data from other innovation surveys in the same country.
Comparability of CIS4 and CIS3 data
The comparability of data between the third and fourth Community Innovation Surveys was improved in comparison with previous surveys due to the fact that they used the same survey methodology, target population, survey questionnaires and definition of innovation.
The CIS 4 questionnaire was shorter and considerably less difficult than the CIS 3 questionnaire. In most countries, CIS 4 was launched in 2005 for the reference period 2004 and an observation period running from 2002 to 2004, while for CIS3 countries used several observation periods.
In order to gain additional information on the innovative capabilities of enterprises, CIS4 asked questions on organisational and marketing innovations and their effects.
Comparability of CIS 2006 and CIS4 data
CIS 2006 largely follows CIS 4 with a few modifications, plus the addition of pilot questions to expand the coverage of marketing and organisational innovation. For CIS 4, an innovative firm is defined as a firm that had introduced a product or process innovation or which had ongoing or abandoned innovative activities. This definition is kept for CIS 2006.
In addition, CIS 2006 has several important simple indicators not included in CIS 4.
An additional breakdown (in addition to NACE and firm size) is the R&D status of the firm (whether or not it performs R&D, with performance defined as either continuous or occasional).
Comparability of CIS 2008 and CIS 2006 data
There are two major changes to the CIS 2008 questionnaire compared to the 2006 version.
First, the CIS 2006 questions on hampering factors and intellectual property rights are not included in the CIS 2008 questionnaire. This is the result of a decision to ask some questions (to which responses change slowly over time) only every four years instead of every two years. The purpose of this decision is to keep the questionnaire short and permit the addition of one-off modules on topics of policy relevance.
Second, the CIS 2008 questionnaire includes a voluntary one-page module on innovation with environmental benefits.
The changes to the CIS 2008 questionnaire were based on the requirement to align with the third revision of the Oslo Manual. Organisational and marketing innovations are now a regular part of the survey. However, fewer questions are asked on organisational and marketing innovation than for product and process innovation.
Comparability of CIS 2010 and CIS 2008 data
CIS 2010 includes the most recent changes recommended by the latest Oslo Manual guidelines: CIS 2010 reports data on product, process, organisation and marketing innovation. The ad-hoc module in the 2010 survey is on the skills available in enterprises and on methods stimulating new ideas and creativity. In this survey wave, the indicators on factors hampering innovation activities are reintroduced. Moreover, the questions about innovation development have changed in that enterprises can tick all types of development they pursued and not only the main one as in CIS 2008. A new innovation activity category for product and process innovations in CIS 2010 is activity to design, improve or change the shape or appearance of goods or services.
15.3. Coherence - cross domain
1- Innovation-related variables are unique to CIS and comparisons with other statistics are not possible.
2- However, the coherence between CIS and SBS statistics is very satisfactory on common variables (turnover, employees).
15.4. Coherence - internal
Not applicable.
Systematic information is not available.
17.1. Data revision - policy
These historical data are no longer updated nor revised.
17.2. Data revision - practice
These historical data are no longer updated nor revised.
18.1. Source data
All aggregations and indicators presented in CIS collections are based on data from the national CIS data collections.
Countries generally carry out a stratified sample survey in order to collect the data, while a number of countries used a census or a mix of census and sample survey. The target population is to be broken down into strata for sampling purposes. The variables to be used for this are size (according to number of employees) and the activity classification (NACE). These two variables are highly correlated with innovation activity. The size-classes should at least be the following 3 breakdowns: 10-49 employees (small), 50-249 employees (medium-sized), 250 + employees (large). Stratification by NACE Rev 2 has been in general by 2-digit level (division) or groups of division until 2010 survey.
The sampling frame to be used for the sample is mostly the official business register.
18.2. Frequency of data collection
The current frequency of CIS is biennial.
18.3. Data collection
Eurostat collects aggregated data and microdata from countries every two years. At the national level, data are mainly collected via mail or online surveys but some countries choose also other collection methods as face to face and phone interviews.
An excel template file is sent by Eurostat protected by password to not allow the countries to change the codes or modify the structure. Once filled by the countries, the file is sent to Eurostat by Edamis and converted into a csv file by Eurostat for production database feed.
18.4. Data validation
The micro and the aggregated data are checked and corrected for detected inconsistencies. Quality controls are done on aggregated and microdata at national level while Eurostat also carries out quality checks. The vast majority of variables are cross-checked to verify the consistency of the answers.
A quality report is produced by the countries in order to report deviations if any.
18.5. Data compilation
At the national level, as in any regular sample survey, the countries extrapolate the collected data with the appropriate weighting scheme for obtaining the population totals. At Eurostat, the aggregates are summed up from the national data.
18.6. Adjustment
If non-response in the survey exceeds 30 % based on the ratio between non-responding and operating enterprises in the sample population as a whole, a non-response analysis is recommended.
At least a 10 % sample of the relevant non-respondents should be drawn and a non-response survey then undertaken. This survey includes the main questions from the ordinary questionnaires in order to distinguish three types of innovative firms:
1. The firm is a product/process (PP) innovator: it introduced a product or process innovation, but did not perform R&D for PP innovation.
2. The firm performs R&D for PP innovation.
3. The firm is only a non-technological innovator: it introduced an organisational innovation, but is not a PP innovator.
In order for the non-response analysis to be useful, the non-response survey should have a very high response rate. This non-response survey should be carried out for at least the core target NACE population. As a final check, the non-response analysis should determine if there is a statistically significant difference in the broad definition of innovation. If that is the case, this information is used to calibrate the weights, using such procedures as CALMAR. Additional auxiliary information can also be incorporated if it is considered that this will enhance the accuracy of the estimates.
In CIS surveys, some countries added a few extra questions for national purposes and some countries excluded some or all of the optional questions.
The Community Innovation Survey (CIS) is a survey of innovation activity in enterprises. The harmonised survey is designed to provide information on the innovativeness of sectors by type of enterprises, on the different types of innovation and on various aspects of the development of an innovation, such as the objectives, the sources of information, the public funding or the expenditures.
The CIS provides statistics broke down by countries, type of innovators, economic activities and size classes. The survey is currently carried out every two years across the EU Member States, some EFTA countries and EU candidate countries.
In order to ensure comparability across countries, Eurostat, in close cooperation with the countries, has developed a standard core questionnaire starting with the CIS3 data collection, along with an accompanying set of definitions and methodological recommendations. The concepts and underlying methodology of the CIS are also based on the Oslo Manual — second edition of 1997 and third edition of 2005 (see link at the bottom of the page).
The data presented in these tables refer to enterprises with ‘10 employees or more’ active in the sectors to be covered under the Regulation (cf. NACE CORE). Further activities may be covered on a voluntary basis. Most statistics are based on a reference period of three years, but some use one calendar year.
Since CIS 2008, the survey has included an ad-hoc module. It consists of a set of questions focusing on a special theme. The themes are different in each survey wave, allowing data to be obtained on specific issues beyond the data usually collected.
Overview over time:
Initially, the CIS data collection was carried out every four years. The first collection (CIS Light) was launched in 1993 as a pilot exercise and the second (CIS2) was carried out in 1997/1998 for most countries except Greece and Ireland, where it was launched in 1999. The third survey (CIS3) was conducted in 2000/2001 for most participating countries with the exception of Norway, Iceland, Luxembourg and Greece, where it was launched in 2002.
As from 2004, the survey has been carried out every two years.
CIS4 was conducted in the EU Member States (as for 2004), Iceland, Norway, Bulgaria and Romania. The survey was launched in 2005 with a three-year reference period 2002 to 2004 for most indicators.
The fifth survey CIS 2006 was carried out in all EU Member States (as for 2006), Norway, Bulgaria, Romania, Croatia and Turkey. It was launched in 2007, mostly for the reference period 2004 to 2006.
As regards CIS 2008, EU Member States (all except Greece), Iceland, Norway, Croatia and Turkey took part in the survey. CIS 2008 was launched in 2009 with a three-year reference period 2006 to 2008 for most indicators. Changes were made to the CIS 2008 questionnaire to bring it into line with the third revision of the Oslo Manual, 2005 edition, by giving greater weight to organisational and marketing innovation. CIS 2008 was complemented by an ad-hoc module on innovation with environmental benefits.
The seventh Community Innovation Survey, CIS 2010, had 31 participating countries (all EU Member States except Greece, Iceland, Norway, Croatia, Serbia and Turkey) and reported most results for the reference period 2008-2010. CIS 2010 also follows the recommendations of the Oslo Manual and reports indicators on four types of innovation: product, process, organisational and marketing.
However, despite implementation of the recommendations of the third edition of the Oslo Manual, the question on innovation expenditures is still limited to product and process innovation in order to maintain continuity with earlier versions of the CIS. Furthermore, generally fewer questions are asked about organisational and marketing innovation than about product and process innovation.
While the European innovation statistics use the aggregated national data, the microdata sets can be accessed by researchers via the SAFE Centre of Eurostat in Luxembourg or via the microdata on CD-ROM releases in more anonymised form; some countries also provide access to their micro-data at similar safe centres.
28 January 2020
CIS provides information on the characteristics of innovation activity at enterprise level. It allows Europe’s progress to be monitored in the area of innovation, creating a better understanding of the innovation process with analysis of the objectives and the effects of innovation. These results can also be linked to variables related to competitiveness, employment and economic growth. The concepts are in line with those recommended by the Oslo Manual (2nd edition 1997 and 3rd edition 2005), which is the internationally recognised standard methodology for collecting innovation statistics.
From CIS3 to CIS2006, the innovation concept was still limited to technological innovation (product and process innovation). The following categories of indicators were provided for these three CIS waves.
Product, process, on-going and abandoned innovation
Innovation activity and expenditure
Intramural research and experimental development (R&D)
Effects of innovation
Public funding of innovation
Innovation cooperation
Sources of information for innovation
Hampered innovation activity
Patents and other protection methods
Other important strategic and organisational changes in the enterprise
From CIS 2008 onwards, information has been collected not only on product and process innovation but also on organisational and marketing innovation. Most questions cover new or significantly improved goods or services or the implementation of new or significantly improved processes, logistics or distribution methods. CIS results offer a broad set of indicators on innovation activities, innovation expenditures, public funding, sources of information for innovation, innovation cooperation, innovation objectives and organisational and marketing innovation.
Main concepts and definitions used for the CIS data collection:
An innovation is the implementation of a new or significantly improved product (good or service), process, new marketing method, or new organisational method in business practices, workplace organisation or external relations. This broad definition of innovation encompasses a wide range of possible innovations. An innovation can be more narrowly categorised as the implementation of one or more types of innovations, for instance product and process innovations. An overview of the types of innovators is available in annex.
The minimum requirement for an innovation is that the product, process, marketing method or organisational method must be new to the firm (or significantly improved). This includes products, processes and methods that firms are the first to develop and those that have been adopted from other firms or organisations.
Product innovative enterprises are those who introduced, during the period under review, new and significantly improved goods and/or services with respect to their fundamental characteristics, technical specifications, incorporated software or other immaterial components, intended uses, or user friendliness. Changes of a solely aesthetic nature and the simple resale of new goods and services purchased from other enterprises are not considered as innovation.
Process innovative enterprises implemented new and significantly improved production technologies or new and significantly improved methods of supplying services and delivering products during the period under review. The outcome of such innovations should be significant with respect to the level of output, quality of products (goods or services) or costs of production and distribution. Purely organisational or managerial changes are not included.
Organisational innovative enterprises implemented a new organisational method in the enterprise’s business practices, workplace organisation or external relations.
Marketing innovative enterprises implemented a new marketing method involving significant changes in product design or packaging, product placement, product promotion or pricing.
Innovative enterprises had innovation activities during the period under review, including enterprises with on-going and abandoned activities. In other words, enterprises that had innovation activities during the period under review, regardless of whether the activity resulted in the implementation of an innovation, are innovation-active.
During a given period, innovation activities can be of three kinds:
successful, in having resulted in the implementation of an innovation (although the innovation need not have been commercially successful);
on-going, with work in progress that has not yet resulted in the implementation of an innovation.
abandoned before the implementation of an innovation.
Non-innovative enterprises had no innovation activity whatsoever during the reference period. These enterprises answered only a limited set of questions from the survey in relation to the absence of innovation activity, factors hampering innovation, patents and other protection methods, etc. The CIS 2008 and CIS 2010 modules (on eco-innovation and on creativity and skills, respectively) targeted enterprises both with and without innovation activity.
The statistical unit is the enterprise, as defined in the Council Regulation on statistical units (Council Regulation (EEC) No 696/93 of 15 March 1993, OJ No L76 of 3 March 1993) or as defined in the statistical business register.
The Regulation defines the enterprise as ‘the smallest combination of legal units that is an organisational unit producing goods or services, which benefits from a certain degree of autonomy in decision making, especially for the allocation of its current resources. It may carry out one or more activities at one or more locations and it may be a combination of legal units, one legal unit or part of a legal unit ’.
The population of the CIS is determined by the size of the enterprise and its principal activity. All enterprises with 10 or more employees in any of the specified sectors are included in the statistical population. For the coverage of economic activities, see section 3.3 (Sector coverage) above.
Most EU Member States took part in CIS Light, CIS 2, CIS 3 and CIS 2004. The number of countries participating has increased over time. Additionally, some EFTA, EU candidate and non-member countries have conducted the survey as well.
CIS2006:
The fifth survey CIS 2006 was carried out in all EU Member States (as for 2006), Norway, Bulgaria, Romania, Croatia and Turkey.
CIS 2008:
The survey was carried out in all EU Member States except Greece. Iceland, Norway, Croatia and Turkey also took part.
CIS 2010
Thirty-one countries participated in the 2010 Community Innovation Survey: all EU Member States (except Greece), Iceland, Norway, Croatia, Serbia and Turkey.
A particular feature of the Community Innovation Survey is that it uses different reference periods (except CIS 2):
most indicators are based on a three-year reference period, for example, the indicator showing whether an enterprise introduced an innovation covers a period of three years;
a minority of indicators are based on one year. It's the case of the turnover and employment indicators, which are usually requested for the first and last years of the 3-year reference period. Innovation expenditure is also based only on the last year of this 3-year period.
CIS Light:
Countries were free to choose between two reference periods of 3 years: 2000-2002 or 2001-2003. Latvia, Slovakia and Norway had data for the period 2001-2003, while most other countries had data for the first period, 2000-2002, with the exception of Luxembourg, Lithuania, Malta and Hungary, which used a 2-year reference period (2002-2003), Slovenia, which used a 2-year reference period (2001-2002), and Portugal, which reported results for only a single reference year (2003).
CIS2
All data referred to the calendar year 1996.
CIS 3
The standard reference period for the indicators with a 3-year period was from 1998 to 2000.
Exceptions to the reference year in CIS3:
Poland generally had a reference period of 1998-2000 for the industrial sector and 1997-1999 for the services sector. The Czech Republic, Hungary, Latvia, Lithuania, Slovakia and Norway had a reference period of 1999-2001. Romania had a reference period of 2000-2002. Slovenia had a two-year reference period of 2001-2002, while Bulgaria had a reference period of 2001-2003. Spain used an earlier version of the CIS3 core questionnaire.
Moreover, there are some exceptions: the data on turnover (tables inn_bas and inn_prod) refer to 2001; the data on expenditure (table inn_exp) are for 2000 for the industry sector and for 1999 for the services sector; the data on non-patent protection methods (table inn_pat) refer to 1996; and the data on non-technological innovation (table inn_cha) refer to 2001.
CIS4
The standard reference period for the indicators with a 3-year period was from 2002 to 2004. Indicators such as innovation expenditures were based on the reference year 2004. The turnover and employment figures were reported for the calendar year 2002 and for the calendar year 2004. All countries collected the data in accordance with these reference periods, with just the Czech Republic having a reference period of 2003-2005.
CIS2006
The standard reference period for the indicators with a 3-year period was from 2004 to 2006. Indicators such as innovation expenditures were based on the reference year 2006, while the turnover and employment figures were reported for the calendar year 2004 and for the calendar year 2006.
CIS 2008
Similar to the previous surveys, three reference periods were used in the questionnaire: the reference period for most questions was 2006 to 2008, i.e. the three-year period from the beginning of 2006 to the end of 2008. The indicators on innovation expenditures were based on the calendar year 2008. Additionally, the turnover and employment of enterprises were requested for two years: 2006 and 2008.
All countries collected the data in accordance with these reference periods.
CIS 2010
Most CIS 2010 questions covered the reference period 2008 to 2010, i.e. the 3-year period from the beginning of 2008 to the end of 2010. As regards the indicators on innovation expenditures for product and process innovations, the reference period was only one year, 2010. Spain also used just 2010 as the reference period to collect the number of enterprises engaged in innovation activities. The other countries used the 3-year period for these indicators. Moreover, a limited number of basic economic indicators such as turnover and employment were requested for 2008 and for 2010.
The overall picture is that all countries make considerable efforts to reduce errors or at least to identify and correct them. Interviewer training and assistance to respondents are provided for data collection. Comprehensive data validation is the norm during and after data collection. Re-contacting enterprises is the primary option for correcting errors, so the use of imputation is less needed. All these factors point to the high quality of the innovation statistics.
CIS data are available in the following units: Absolute value (NBR) and Percentage (PC).
All the financial variables (innovation expenditures and turnover) are presented in thousand euros.
At the national level, as in any regular sample survey, the countries extrapolate the collected data with the appropriate weighting scheme for obtaining the population totals. At Eurostat, the aggregates are summed up from the national data.
All aggregations and indicators presented in CIS collections are based on data from the national CIS data collections.
Countries generally carry out a stratified sample survey in order to collect the data, while a number of countries used a census or a mix of census and sample survey. The target population is to be broken down into strata for sampling purposes. The variables to be used for this are size (according to number of employees) and the activity classification (NACE). These two variables are highly correlated with innovation activity. The size-classes should at least be the following 3 breakdowns: 10-49 employees (small), 50-249 employees (medium-sized), 250 + employees (large). Stratification by NACE Rev 2 has been in general by 2-digit level (division) or groups of division until 2010 survey.
The sampling frame to be used for the sample is mostly the official business register.
Since 2004, Community Innovation Surveys have been conducted every two years.
According to Commission Regulation No 1450/2004, national CIS statistics must be delivered to Eurostat within 18 months from the end of the reference year.
Most countries conform to the Regulation and delivered the data by the requested time.
The CIS2 data were released in the first quarter of 1999. The CIS3 data were disseminated by Eurostat in August 2003. The CIS4 data were disseminated in October 2006. The CIS 2006 data collection was disseminated in October 2008. The CIS 2008 data collection was released in November 2010. Eurostat made available CIS 2010 data at the beginning of November 2012.
In order to ensure comparability across countries, Eurostat, in close cooperation with the EU Member States and other countries, has developed a standard core questionnaire for each CIS round, with an accompanying set of definitions and methodological recommendations.
Comparability of CIS3 and CIS2 data
Comparability of data between the second and third Community Innovation Surveys is limited due to differences in survey methodology, target population, survey questionnaires and the definition of innovation. These differences concern:
Lower cut-off for inclusion in the target population
The cut-off for inclusion in the target population in CIS3 is firms with 10 employees and more in both the manufacturing and service sectors. In CIS2 the cut-off for inclusion in the target population is firms with 20 employees in the manufacturing sector and firms with 10 employees in the service sector.
More industries included in the target population
The CIS3 target population includes more activities defined by NACE than the CIS 2 survey. The additional activities covered by CIS3 are mining and quarrying (NACE 10-14), research and development (NACE 73) and technical testing and analysis (NACE 74.3). In addition, all NACE activities 63-64 (storage and communication) and the entire NACE activity 74.2 (architectural and engineering activities) were included.
Changed definition of innovation
The definition of innovation changed slightly in CIS3 compared to CIS2. In CIS3 the term ‘Innovation’ is used instead of ‘Technological innovation’. The term ‘technological’ remains though in the explanatory text of the concept.
Same core questionnaire used for manufacturing and services enterprises
In CIS3 a common core questionnaire for both kinds of enterprise (manufacturing and services) was used. This was not the case with CIS2, where two different core questionnaires were used for these two sectors. Enterprises in the service sector were asked approximately 90 per cent of the questions of the core questionnaire for enterprises in manufacturing. In the core questionnaire for service enterprises in CIS2 there was no split between product and process innovations, and no questions were asked on turnover from products new to the enterprise or turnover from products new to the market. The use of one common core questionnaire for CIS3 provides more harmonised data on overall innovation activities. The data are also more comparable between the two main sectors.
More information collected on both innovators and non-innovators
The CIS3 core questionnaire is substantially larger than the two core questionnaires used in CIS2. It therefore provides a more complete picture of both innovating and non-innovating enterprises. Under the assumption that the reporting enterprise was both product- and process-innovative, the CIS3 core questionnaire is approximately 50 per cent bigger than the CIS2 questionnaire in terms of the number of questions. More information is also collected on non-innovative enterprises in the CIS3 core questionnaire since non-innovative firms are asked more questions in CIS3 than in CIS2.
Both CIS Light and CIS2 were implemented on a voluntary basis by countries. Therefore no standardised methodology or questionnaire was implemented at national level. The differences in data collection may affect both inter-country and intra-country comparison, i.e. data may not be comparable either between countries or between CIS and data from other innovation surveys in the same country.
Comparability of CIS4 and CIS3 data
The comparability of data between the third and fourth Community Innovation Surveys was improved in comparison with previous surveys due to the fact that they used the same survey methodology, target population, survey questionnaires and definition of innovation.
The CIS 4 questionnaire was shorter and considerably less difficult than the CIS 3 questionnaire. In most countries, CIS 4 was launched in 2005 for the reference period 2004 and an observation period running from 2002 to 2004, while for CIS3 countries used several observation periods.
In order to gain additional information on the innovative capabilities of enterprises, CIS4 asked questions on organisational and marketing innovations and their effects.
Comparability of CIS 2006 and CIS4 data
CIS 2006 largely follows CIS 4 with a few modifications, plus the addition of pilot questions to expand the coverage of marketing and organisational innovation. For CIS 4, an innovative firm is defined as a firm that had introduced a product or process innovation or which had ongoing or abandoned innovative activities. This definition is kept for CIS 2006.
In addition, CIS 2006 has several important simple indicators not included in CIS 4.
An additional breakdown (in addition to NACE and firm size) is the R&D status of the firm (whether or not it performs R&D, with performance defined as either continuous or occasional).
Comparability of CIS 2008 and CIS 2006 data
There are two major changes to the CIS 2008 questionnaire compared to the 2006 version.
First, the CIS 2006 questions on hampering factors and intellectual property rights are not included in the CIS 2008 questionnaire. This is the result of a decision to ask some questions (to which responses change slowly over time) only every four years instead of every two years. The purpose of this decision is to keep the questionnaire short and permit the addition of one-off modules on topics of policy relevance.
Second, the CIS 2008 questionnaire includes a voluntary one-page module on innovation with environmental benefits.
The changes to the CIS 2008 questionnaire were based on the requirement to align with the third revision of the Oslo Manual. Organisational and marketing innovations are now a regular part of the survey. However, fewer questions are asked on organisational and marketing innovation than for product and process innovation.
Comparability of CIS 2010 and CIS 2008 data
CIS 2010 includes the most recent changes recommended by the latest Oslo Manual guidelines: CIS 2010 reports data on product, process, organisation and marketing innovation. The ad-hoc module in the 2010 survey is on the skills available in enterprises and on methods stimulating new ideas and creativity. In this survey wave, the indicators on factors hampering innovation activities are reintroduced. Moreover, the questions about innovation development have changed in that enterprises can tick all types of development they pursued and not only the main one as in CIS 2008. A new innovation activity category for product and process innovations in CIS 2010 is activity to design, improve or change the shape or appearance of goods or services.
