7.1. Indicators

The European Innovation Scoreboard (EIS) covers the 27 EU Member States, Croatia and Turkey, the associate countries Iceland, Norway and Switzerland, as well as Australia, Canada, Israel, Japan and the US. The indicators of the EIS summarise the main elements of innovation performance.

In 2005, the EIS has been revised in collaboration with the Joint Research Centre [1]. The number of categories of indicators was increased from four to five and the set of innovation indicators was modified and increased to 26. The EIS 2005 Methodology Report (MR) (available on the INNO Metrics website [2]) describes and explains all changes in full detail. The EIS 2006 implemented three changes. The indicator measuring the share of university R&D expenditures financed by the business sector was removed; the indicator on public R&D expenditures, which was defined as the differences between total R&D expenditures and business R&D expenditures, was redefined as the sum of government R&D expenditures and university R&D expenditures only; and the indicator on the share of SMEs using non-technological change was changed into the share of SMEs using organisational innovation following the change in the survey questions on non-technological change from the third Community Innovation Survey (CIS-3) to the fourth Community Innovation Survey (CIS-4).

The EIS 2007 fully implements the list of indicators from the EIS 2006. The innovation indicators are assigned to five dimensions and grouped in two main themes: inputs and outputs. Table 5 shows the 5 main categories, the 25 indicators [3], and the primary data sources for each indicator [4]. Innovation inputs cover three innovation dimensions: Innovation drivers measure the structural conditions required for innovation potential; Knowledge creation measures the investments in R&D activities, considered as key elements for a successful knowledge-based economy; and Innovation & entrepreneurship measures the efforts towards innovation at firm level. Innovation outputs cover two innovation dimensions: Applications measures the performance, expressed in terms of labour and business activities, and their value added in innovative sectors; and Intellectual property measures the achieved results in terms of successful know-how.

Table 5: EIS 2007 Indicators

OHIM: Office of Harmonization for the Internal Market

7.2. Methodology of calculating the Summary Innovation Index

The SII 2007 is calculated as follows:

1. Calculate for every indicator and for every country the most recent relative to the EU score. E.g. if for country A the most recent data point is 500 for year 2005, for country B 400 for year 2004, and the EU scores for 2004 and 2005 are respectively 100 and 125, then the relative to EU score for country A is 100*(500/125)=400 and for country B 100*(400/100)=400. By calculating relative to EU scores business cycles effects will be minimized when timeliness of data availability differs between countries (cf. Annex B for differences in most recent years between countries). Possible outliers are identified as those scores which are higher than the EU average plus 3 times the standard deviation. These outliers are not included determining the maximum relative to EU scores.
2. Calculate re-scaled scores of the indicator data by first subtracting the lowest value found within the group of EU27 countries, Iceland, Norway and Switzerland (thus excluding non-European countries and European countries where data availability is less than 75%) and then dividing by the difference between the highest and lowest values found within the group of EU27 countries, Iceland, Norway and Switzerland. The maximum re-scaled score is thus equal to 1 and the minimum value is equal to 0. For Croatia, Turkey, Australia, Canada, Israel, Japan and the US for those cases where the value of an indicator is above the maximum relative to EU score or below the minimum relative to EU score the re-scaled score is set equal to 1 respectively 0. Countries where indicator scores were identified as a possible outlier (cf. Step 1) receive a re-scaled score of 1.
3. The SII 2007 is then calculated as the average value of all re-scaled scores where indicators for which data are available receive the same weight. The SII is by definition between 0 and 1 for all countries.

For the CIS indicators EU mean values are available from Eurostat. EU mean scores are calculated separately for each CIS indicator dividing the sum of all numerator data for those countries for which CIS data are available by the sum of all denominator data. In fact, as only CIS-4 data are used, these EU mean values are not necessary for calculating the re-scaled indicator scores but they illustrative purposes as shown in the relative to EU performance charts for each country.
The SII values for those countries where data is missing for 8 or more indicators – Croatia, Turkey, Australia, Canada, Israel, Japan and the US – are estimated as follows:

1. Calculate for all countries a summary innovation index using only data for the 18 non-CIS indicators (“non-CIS SII”).
2. For the EU27 countries, Iceland, Norway and Switzerland a simple linear regression is performed with the “non-CIS SII” as the dependent variable and the SII as the independent variable. The estimated regression coefficient equals 1.0742, the estimated constant -0.0478 and the R2 equals 0.950. The regression coefficients are significant at the 1% level and 5% level respectively.
3. For Australia, Croatia, Canada, Japan, Israel, Turkey and the US the SII 2007 is then calculated by dividing the difference between the “non-CIS SII” and the value for the estimated constant by the value for estimated regression coefficient: SII 2007 = (“non-CIS SII” – (-0.0478)) / 1.0742.

7.3. Methodology of calculating the SII growth rate

The SII growth rate is based on SII values over a 5-year period. These SII values are calculated differently than the SII 2007 as we use maximum and minimum scores of the full 5 years (denoted as T-4, T-3, T-2, T-1 and T, where T comes closest to the years used for calculating the SII 2007) so the SII scores will also identify changes in improvement for those countries showing highest performance in individual indicators.
The procedure is as follows:

1. Calculate for every indicator and for every country the relative to EU scores (cf. Step 1 above).
2. Most recent data are then used for year T etc. If data for a year-in-between is not available we substitute with the value for the next year. If data are not available for all 5 years, we replace missing values with the latest available year. Two examples will clarify this step.

1. Calculate re-scaled scores of the indicator data by first subtracting the lowest value found for all 5 years within the group of EU27 countries, Iceland, Norway and Switzerland and then dividing by the difference between the highest and lowest values found for all 5 years within the group of EU27 countries, Iceland, Norway and Switzerland. The maximum re-scaled score is thus equal to 1 and the minimum value is equal to 0. For Croatia, Turkey, Australia, Canada, Israel, Japan and the US for those cases where the value of an indicator is above the maximum relative to EU score or below the minimum relative to EU score the re-scaled score is set equal to 1 respectively 0. Note that these scores can differ from those calculate under Step 1 if either the maximum or minimum value within the group of EU27 countries, Iceland, Norway and Switzerland is found for a year prior to the most recent year.
2. The SII scores are then calculated as the average value of all re-scaled scores where indicators for which data are available receive the same weight.

For the CIS indicators the CIS-4 results are used for all 5 years. The SII values for those countries where data is missing for 8 or more indicators – Croatia, Turkey, Australia, Canada, Israel, Japan and the US – are estimated for each year using the procedure as outlined in Steps 4 to 6 above.
The growth rate of the SII is then calculated as the annual percentage change between the SII in year T and the average over the preceding three years, after a one-year lag (i.e. T-4, T-3 and T-2). The three-year average is used to reduce year-to-year variability; the one-year lag is used to increase the difference between the average for the three base years and the final year and to minimize the problem of statistical/sampling variability.

7.4. Calculation of time to convergence

The time to convergence can be calculating using a linear and non-linear approach. The linear approach assumes a simple extrapolation of the current SII trend rate:

 is the growth rate of the SII for country X and  equals the SII 2007 at time T. The SII for country X at time T equals the current SII for country X multiplied by the current SII growth rate to the power T.
The non-linear approach takes into account that it will become more and more difficult to maintain high growth rates. The non-linear approach assumes that the growth rate of each country will diminish over time with the rate of decrease depending on the size of the initial gap (i.e. the larger the initial gap, the faster the subsequent rate of decline):

The SII for country X at time T equals the SII of the previous year for country X multiplied by a reduced version of the SII growth rate where the size of the reduction depends on the initial gap with the EU and decreases over time with a diminishing rate of decrease.