1 Problem & Solution
1.1 Few Primary Data Sources
1.1.1 Poor Representation in SNA: No Value Added Information
The SNA is the most important source of system where innovation data is measured or related to. The most detailed measurement in the SNA is made within institutional units, and far less attention is paid to the households and NPISHs, and even the government sector. Most innovation data is collected from the business enterprise sector.
The Oslo Manual’s guidelines are not expressly designed to measure innovation in other SNA sectors, but research shows that many of the concepts can be applied to them (Gault, 2018). Oslo Manual p59
The most important is the lack of large and medium sized businesses, and the dominance of microenterprises. These enterprises file simplified tax returns and financial statements – which are the data sources of value added. They are also exempted from almost all statistical data gathering, including innovation due to their size.
1.1.2 Atypical Labor Arragenments, No Employment Data
- They usually employ up to 2 people, and the main form of employment is atypical, freelance, which is not well represented in employment-based data collection.
1.1.3 Not Suitable Classification And Aggregation
- The CCIs are parts of larger NACE classification groups, so no statistics are made for music, film, design, etc.
1.1.4 Poor Representation in CIS
The main source of the European and OECD innovation information is the CIS survey that is designed on the basis of the Oslo Manual. The CIS survey only coverst enterprises with at least 10 employees, therefore excempting almost all CCIs. In many EU members states and industries, there may be not a single representative of a cultural and creative industry in the survey.
1.2 Disaggregation Strategies
We have dealt with these issues in the following way in the music and the film industry. - We have created microenterprise surveys in the music and film industry to establish variable values that allow to “pull out” the music and the film industry from larger categories on a factual basis.
- With accountants and tax advisors we have reviewed the actual accounting of film projects to understand the structure of inputs and outputs and be able to factually “pull out” the film industry from its larger NACE group and place it as a new row and column in the SIOT which allows the calculation of direct, indirect and multiplier effects. For example, a country that had a standard 60x60 SIOT table for such calculations became a 61x61 table extended with the film industry (and the film industry subtracted from the original containing row and column.)
1.3 Filling the Data Gaps
Subject- versus object-based approaches
The most common use of the object-based approach is to collect data on specific innovations, for example innovations reported in trade journals, crowdfunding platforms or, in a survey context, the most important innovation for a given organisation
The subject approach is commonly used in innovation surveys to collect data on the innovation activities, outputs and outcomes of the respondent’s organisation. Subject-based surveys can benefit from the statistical infrastructure of business registers and other available information at the firm level, including the industry of activity and the number of employees.
- We included harmonized questions of interest in these surveys. For example, one can think of a music or design industry questionnaire that fills out the missing Eurostat-ESSNet-government gaps, and also the missing questions of the CIS survey on innovation (which excludes microenterprises and most small enterprises.)
1.3.1 Data Linkage
Data Linkage means that a survey is designed to fill out the missing gaps from publicly available information about a sample of the target population. This methodology is often used in innovation surveys which target enterprises with at least 10 employees. As the questinnaire is sent to a legal person and not natural persons, company registry data or other public information can be linked with a survey.
It should be examined to what extent can this strategy reduce the data collection burden on CCIs. Most CCIs fall under the microenterprise category, and their publicly available data is very limited (they create simplified financal reports.) Furthermore, a very large proportion are not legal persons, and fall under natural data protection rules, making the data linkage very difficult.
1.3.2 Ex Ante Survey Harmonization
A logical strategy is to re-create existing data sources of innovation data for the CCIs with a simplified, ex ante harmonized CCI-CIS survey that is applicable for microenterprises with few or no employees, and enterprises without legal personality.
1.3.3 Survey Harmonization
1.3.4 Alternative Data Sources
Because most CCIs as institutions are excempted from the data collection systems, we may try to look for surveys targeting individuals. For example,
1.4 Re-Weigthing and Post-Stratification
The Oslo Manual recognizes that random sampling of the target population of innovation surveys is an ineffective method, and recognizes the need for stratified sampling. We have developed ways to create well-stratified samples of music professionals connecting anonymized collective management payout data with anonymous natural person surveys.
Another strategy may be the re-weighting, re-balancing of the extremely skewed but existing primary data. For example, the CIS surveys may containg a few CCIs, but they are not representative because of the extremely skewed sample due to the employee number limit. One may be hinted to try to re-weight existing survey data to represent the true distribution of CCI data, for example, using a smaller auxilliary survey for this purpose.
This approach is a very risky approach because of the extreme skew. For example, it is likely that in some countries the only music industry player meeting the size threshold may the national collective management society, which is a very special and non-representative member of the music industry.
The main information system of economic statistics is the system of national accounts (SNA). The EU uses so called Symmetric Input-Output Tables (SIOTs) in the SNA system to make economic impact assessments, multiplier, direct and indirect effect analysis. The SIOTs do not directly contain the CCIs (because they are hidden in larger NACE classes) and even these data is estimated, because these NACE classes are dominated by microenterprises, whose employment, value added and innovation data is missing. The Eurostat-ESSnet network suggests member states to carry out representative surveys of microenterprises who are omitted from the mandatory and comprehensive data collection, but very few member states do this, and even if they do, they do not harmonize among each other these efforts.