Understanding R&D Internationalization through Firm and Patent Data Analysis

Explore the measurement of R&D internationalization by analyzing firm and patent data. Traditional and recent approaches, including combining firm and patent data, offer insights into foreign R&D activities. The International Hierarchical Ownership Model (IHOM) provides a longitudinal view of MNC ownership hierarchies, detecting changes over time.

• R&D
• Internationalization
• Firm Data
• Patent Analysis
• IHOM

ztya Follow

Uploaded on Sep 15, 2024 | 0 Views

Download Presentation

Please find below an Image/Link to download the presentation.

The content on the website is provided AS IS for your information and personal use only. It may not be sold, licensed, or shared on other websites without obtaining consent from the author. Download presentation by click this link. If you encounter any issues during the download, it is possible that the publisher has removed the file from their server.

Presentation Transcript

1. Measuring the extent and effects of R&D internationalization by matching firm and patent data Bj rn Jindra, PhD Copenhagen Business School Universit t Bremen 4th INNOVA MEASURE III Expert Workshop 27.11.2018, JRC Ispra, Italy

2. Traditional approaches 2

3. Using R&D statistics Based on company surveys and/or company reports Different approaches to measure foreign R&D across countries Differentiation by home industry, firm size etc. Aggregation at group level possible (ownership structure) Limited data on host countries/industries for foreign R&D Using patent based indicators Triade/international patents Cross-border approach Patent classes and technological areas Lack of information on ownership 3

4. Recent approaches combining firm and patent data IPTS-OECD-pilot* (Dernis et al. 2015, R&D Scoreboard 2015/16, Daiko et al 2017) Matching patent applications at 5 IP offices to 2.000 top R&D investors worldwide Initial results 25% of inventors are located abroad In selected sectors >50% (including Pharma, Finance, Mining) Advantages Data on location of foreign inventors and technological areas Ownership structures taken into account Limitation Static ownership structure (end of 2012) Short observation period for patents (3 years) Note: *Similar approach by Laurens et al (2014) 4

5. Creating dynamic ownership structures in large datasets 1. Starting population (Source: BvD Orbis online, 2014) Firms located in EU country & at least one foreign subsidiary Primary NACEr2-code in section C & min. 100 employees 2. Hierarchical Ownership Model (Source: BvD Orbis) 1,632 MNEs located in EU15 (domestic ultimate owner, industrial enterprise at the top of the MNE hierarchy) all foreign/domestic entries/exist (2007-2014) 3. Matching with patent data (Source: BvD Orbis/PATSTAT) Identification of patent documents of parents/subsidiaries Identification of 350.000 priority patent applications (2007-2013) Origin of inventor analysis (fractional counting) Consolidation of patent data at MNC level 5

6. The International Hierarchical Ownership Model (IHOM) 6

7. What is the IHOM? A longitudinal model of the MNC ownership hierarchies Detects changes in MNC networks over time Detects entries, exits, ownership level changes, etc. 7

8. Data extraction Recursive Hierarchy Extraction Process For each year 5.1 million Ownership Links Extracted 8

9. Main issues in extraction process (I) The Issue: extraction based on currently active entities (2014) excludes dissolved or divested entities (e.g. exits) The solution: use all extracted historical version ids in union with current versions! 9

10. Main issues in extraction process (II) The issue: Data gaps, duplicates, inconsistencies, etc. Interpolation prevents fake exits and entries Selection prevents invalid count based measures 10

11. Building hierarchies For each MNC and year a downward hierarchal search was performed and total MNC ownership of each entity was calculated Example: Siemens AG 2010 1235 Subsidiaries (>25%) 1023 Foreign 213 Domestic 1776 Ownership paths 11

12. Identifying true MNC Parents The initial sample was corrected to include alone actual European MNCs Some sample firms were owned by other firms Some were large subsidiaries of non-European MNCs The solution: The IHOM catered for both cases Carlsberg Breweries owned by Carlsberg A/S (100%) KIA Motors Europe Gmbh (DE) Directly owned by KIA Motors Corporation (KR) Included in the sample although a regional headquarters of KIA Both are large European entities with foreign subsidiaries and was included in the initial sample 12

13. Identifying Entries Two ways of identifying entries: 1. Greenfield by date of incorporation 2. Acquisition by ownership change To ensure high validity records documenting prior ownership was required for acquisition entries Prevents fake entries in the beginning of the analysis period 13

14. Identifying Exits Two ways of identifying exits: 1. Dissolve by date of dissolve 2. Acquisition by ownership change To ensure high validity records documenting subsequent ownership was required for acquisition exits Prevents fake exits in the end of the analysis period 14

15. Preliminary descriptive results - EU 15 sample - 15

16. EU15 sample Entries/exits Number of subsidiaries Number of subsidiaries Entries Exits 2007 no. 24,710 32.00 18,101 52,504 68.00 51,785 2007 - 2014 no. 2007 - 2014 no. 13,068 32,173 2014 % % % no. % Domestic Foreign 25.90 74.10 28.89 71.11 29,743 72,116 29.20 70.80 Total 77,214 100 69,886 100 45,241 100 101,859 100 Source: BvD Orbis 2007-2014. Calculation by authors. 16

17. EU15 sample Subsidiary Location Number of subsidiaries 2007 Entries Exits Number of subsidiaries 2014 Change 2007-2014 2007-2014 2007 - 2014 50,634 65.58 38,776 55.48 28,998 64.10 60,412 59.31 9,778 -6.26 EU15 4,598 5.95 6,103 8.73 2,757 6.09 7,944 7.80 3,346 1.84 BRICS 10,525 13.63 11,953 17.10 6,519 14.41 15,959 15.67 5,434 2.04 West* 11,457 14.84 13,054 18.68 6,967 15.40 17,544 17.22 6,087 2.39 RoW 77,214 100 69,886 100 4,5241 100 101,859 100 - - Total Note: *West refers to US, JP, CA, AU, NO, CH Quelle: BvD Orbis 2007-2014. Calculation by authors. 17

18. EU15 MNE priority patent applications (07-13) % applications by subsidiaries % applications with foreign inventors (outside EU15) 80 60,000 70 68.6 68.3 67.1 55,000 66.6 65.7 62.8 60 59.6 50,000 50 40 45,000 30 27.3 40,000 26.6 26.3 24.0 23.6 22.7 22.4 20 35,000 10 0 30,000 2007 2008 2009 2010 2011 2012 2013 Number of MNE priority patent applications (left axis) Share of applications by subsidiaries in % (right axis) Share of MNE priority applications with outside EU15 inventor" 18 Source: BvD Orbis 2016 & PASTAT 2015. Own calculations.

19. EU15 sample Location of inventors 2007 2008 2009 2010 2011 2012 2013 Change % % % % % % % EU15 73.73 72.67 73.40 77.65 77.33 76.00 76.44 2.7 BRICS 2.15 2.72 3.62 3.47 4.42 5.53 6.24 4.1 West* 16.15 18.17 15.79 11.59 11.50 11.93 12.09 -4.1 Rest 7.97 6.44 7.19 7.29 6.75 6.54 5.22 -2.8 Total Note: *West refers to US, JP, CA, AU, NO, CH 100 100 100 100 100 100 100 Source: BvD Orbis 2016 & PASTAT 2015. Own calculations. 19

20. Preliminary descriptive results - German subsample - 20

21. German MNE priority patent applications (07-13) % applications by subsidiaries 70 29,000 68 28,000 65 65 65 27,000 61 60 59 26,000 56 55 53 25,000 50 24,000 45 23,000 40 22,000 35 21,000 30 20,000 2007 2008 2009 2010 2011 2012 2013 Anzahl Priorit tsanmeldungen MNUs (linke Achse) Anteil Anmeldungen durch T chterunternehmen in % (re. Achse) 21 Source: BvD Orbis 2016 & PASTAT 2015. Own calculations.

22. Distribution of subsidiaries with at least 1 priority patent application 100% 7 8 9 9 9 9 11 3 2 3 90% 4 3 4 4 8 7 7 6 7 7 8 80% 11 11 10 12 12 11 12 70% Andere 60% 27 28 28 28 EU13 28 30 27 50% BRICS 40% West 30% EU14 47 46 46 45 43 DE 43 42 20% 10% 0% 2007 2008 2009 2010 2011 2012 2013 Total Total 1,250 subsidiaries 1,237 Subsidairies 22

23. Share of priority patent applications with foreign inventors (by type of applicant) 60 56 55 54 53 53 51 50 50 40 40 38 37 36 36 35 34 30 20 14 13 12 10 9 8 8 8 0 2007 2008 2009 2010 2011 2012 2013 MNUs gesamt M tter T chter 23

24. Geographic distribution of foreign inventors by German MNEs 100% 2 2 2 1 1 1 2 5 6 90% 7 8 11 12 11 80% 27 38 70% 34 30 27 39 28 Andere 60% EU13 50% BRICS West 40% EU14 30% 58 55 54 54 52 51 48 20% 10% 0% 2007 2008 2009 2010 2011 2012 2013 24 Source: BvD Orbis 2016 & PASTAT 2015. Own calculations.

25. Top 15-Inventor countries for German MNEs Top 15 Top 15 2007 2007 15.982 3.045 2.351 764 424 417 347 306 268 262 147 Top Top 15 2013 2013 14.912 1.940 1.749 681 553 513 471 395 262 206 190 186 15 Changes Changes 2007 2007 2013 Top-5-WINNERS 2013 DE JP FR GB AT ES CN KR US FI IT DK CH SE PT 63,87 12,17 9,40 3,05 1,69 1,67 1,39 1,22 1,07 1,05 0,59 0,32 0,26 0,24 0,24 DE FR JP GB CN AT US FI ES IN KR DK IT CH NL 65,32 8,50 7,66 2,98 2,42 2,25 2,06 1,73 1,15 0,90 0,83 0,82 0,31 0,29 0,27 CN US IN FI AT DK RU 206 203 183 133 89 105 37 +1,04 +0,99 +0,81 +0,68 +0,55 +0,49 +0,17 Top-5-LOOSERS -1.295 -410 -155 -116 JP FR ES KR IT PT -4,50 -0,90 -0,52 -0,39 -0,27 -0,21 81 65 60 59 70 67 63 -76 -52 25 Source: BvD Orbis 2016 & PASTAT 2015. Own calculations.

26. German MNEs Locations for production and inventions (07-14) Production Production 2007 2007 Inventions Inventions 2007 2007 2014 2014 2013 2013 No. Subs. 4.026 26,6 7.017 5.057 33,4 6.144 704 4,7 BRICS 974 West* West* 2.687 17,8 4.762 Rest Rest 1.671 11,1 2.877 No. Subs. No. Patents 15.982 4.623 108 428 3.425 457 No. Patents % % 29,5 25,9 5,2 7,3 20,0 12,1 % 63,9 14.912 65,3 18,5 4.245 0,4 146 1,7 861 13,7 2.359 1,8 307 % DE DE EU14 EU14 EU13 EU13 BRICS 18,6 0,6 3,8 10,3 1,3 1.235 1.724 6,4 15.119 100 23.759 100 25.023 100 22.829 100 Quelle: Source: BvD Orbis 2016 & PASTAT 2015. Own calculations. *US, JP, CA, AU, NO, CH. 26

27. Top-10 Industries with most priority patent applications for German MNEs (07-13) No. Patents 36.210 % foreign inventors 56 NACE 2910 2811 Primary code Manufacture of motor vehicles Manufacture of engines and turbines, except aircraft, vehicle and cycle engines Manufacture of electronic components Manufacture of other chemical products n.e.c. Manufacture of footwear Manufacture of basic iron and steel and of ferro-alloys Manufacture of industrial gases Manufacture of pharmaceutical preparations Manufacture of soap and detergents, cleaning and polishing preparations Cold drawing of wire 18.577 1.891 858 757 595 583 447 36 26 18 95 22 25 19 2611 2059 1520 2410 2011 2120 2041 2434 378 350 15 81 Source: BvD Orbis 2016 & PASTAT 2015. Own calculations. 27

28. Top 5-Industries with most priority patent applications with inventors from China (07-13) Share of patents in the industry No. Patents NACE 2811 Primary code Manufacture of engines and turbines, except aircraft, vehicle and cycle engines Manufacture of motor vehicles Manufacture of other chemical products n.e.c. Manufacture of other chemical products n.e.c. Manufacture of pharmaceutical preparations 2.974 626 186 74 60 5,75 0,97 3,94 2,95 2,50 2910 2059 2041 2120 Source: BvD Orbis 2016 & PASTAT 2015. Own calculations. 28

29. Further Research opportunities 29

30. Whats in it? New descriptive insights Changing dynamics in firms location of production/inventions Changing nature of corporate technological activities Nature of European activities in China Firm specific drivers of R&D internationalisation Specialisation & related/unrelated technological diversification Role of acquisitions and divestments (domestic and foreign) Effects of R&D internationalization at the firm level Inventions (and R&D-employment) at home Firm performance at home Extensive/intensive margin 30

31. Limitations and next steps Some of them well known drawbacks of BvD data Limited transparency on usage/change of data providers Expansion of coverage over time Availability and quality of data varies for countries Limited information on representativeness of firm data per country Missing values (in particular at subsidiary level) Moving time window (9 years) for ownership histories Next steps How can we improve the quality of R&D data (domestic, foreign, group) in the dataset? Quality of patents and consider IP bundles Possible extension/update (2007-2018) 31

32. References EU Industrial R&D Scoreboard (2015) The 2015 EU Industrial R&D Investment Scoreboard, Seville (ES): European Commission - Joint Research Centre, Institute for Prospective Technological Studies. EU Industrial R&D Scoreboard (2016) The 2016 EU Industrial R&D Investment Scoreboard, Seville (ES): European Commission - Joint Research Centre, Institute for Prospective Technological Studies. Daiko T., Dernis H., Dosso M., Gkotsis P., Squicciarini M., Vezzani A. (2017) World Corporate Top R&D Investors: Industrial Property Strategies in the Digital Economy. A JRC and OECD common report. Luxembourg: Publications Office of the European Union. Dernis, H., Dosso, M., Herva s, F., Millot, V., Squicciarini, M., Vezzani, A. (2015) World Corporate Top R&D Investors: Innovation and IP bundles. A JRC and OECD common report. Luxembourg: Publications Office of the European Union. Laurens P., Le Bas C., Schoen A., Villard L., Laredo P., 2014. The rate and motives of the internationalization of large firm R&D: toward a turning point? Presented at the International Workshop on the Geography of Innovation, Utrecht, http://elio6.eurolio.eu/indico/ World 32

33. Annex 33

34. A1: Home country distribution EU 15 sample mnc_country AT BE DE DK ES FI FR GB GR IE IT LU NL PT SE Total Freq. Percent Cum. 21 62 324 39 135 88 103 343 20 17 331 1,29 3,80 19,85 2,39 8,27 5,39 6,31 21,02 1,23 1,04 20,28 0,12 3,98 0,98 4,04 1,29 5,09 24,94 27,33 35,60 40,99 47,30 68,32 69,55 70,59 90,87 90,99 94,98 95,96 100,00 2 65 16 66 1,632 100 34