Skip to main content
Log in

Which travels farther? Knowledge or rivalry?

  • Original Paper
  • Published:
The Annals of Regional Science Aims and scope Submit manuscript

Abstract

Bringing firms together generates knowledge spillovers which in turn creates innovation and economic growth. However, firm interactions also generate market rivalry. Unlike knowledge spillovers, we know little over the geographic reach of market rivalry. The paper aims to provide a rigorous comparison of these two channels in a large scale. It approximates knowledge spillovers with the staple metric of patent citations. For market rivalry, the metric of trademark oppositions is proposed. The baseline analysis is at the EUIPO, for trademark oppositions, and EPO for patent citations. The country level and NUTS-3 analyses show that there is stronger home bias for knowledge spillovers compared to market rivalry. To provide robustness, similar data are computed for the USPTO, the most populous patent and trademark office in the world, and an office with substantially fewer trademark oppositions due to its different procedures. State- and county-level analyses provide similar results to the baseline analysis.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2

Similar content being viewed by others

Notes

  1. Such knowledge flows could take place with both parties’ consent formally or informally or without. In the latter case, knowledge flows could be noted as knowledge spillovers. The terms knowledge flows and spillovers are used interchangeably throughout the text.

  2. Firm-level studies have also found that competition can moderate innovation (Bloom et al. 2013; Lychagin et al. 2016; Lucking et al. 2018).

  3. Further, in early studies, Schmoch (2003) and Mendonça et al. (2004) showed that trademarks can to an extent also approximate innovative activity. Later studies have examined a multitude of dimensions with respect to trademarks. For instance, new trademarks can increase firm value (Castaldi and Giarratana 2018), while Mendonça (2014) showed their contribution at the country level. For literature reviews, see Schautschick and Greenhalgh (2016), Castaldi and Dosso (2018) and Castaldi (2019).

  4. For this reason, trademarks registered at this office are often referred to as Community trademarks.

  5. For a meta-analysis of studies that test theories of regional growth, including Porter’s see De Groot et al. (2016).

  6. Feldman and Audretsch (1999) state: “It should be emphasized that by local competition, Jacobs (1969) does not mean competition within product markets… Rather, Jacobs is referring to the competition for the new ideas embodied in economic agents.”.

  7. While studies have employed the Herfindahl index to approximate rivalry, the latter can be present in industries with vastly different concentration sizes (Porter 1990).

  8. In the European Union (EU), the Office for Harmonisation in the Internal Market (OHIM) was founded in 1994 with purpose of accepting and registering, after due examination, trademark applications (Council Regulation (EC) No 40/94). In 2016, OHIM changed its name to European Union Intellectual Property Office (EUIPO).

  9. EC No 207/2009, Article 38. In the latest amendment of the regulation (EC No 2017/1001, Article 43), entities have the option to stop receiving these notifications. However, this is outside the sample period and thus not influencing any results.

  10. von Graevenitz’s study (2009) offers some additional support for this. He shows that when firms are very close competitors, then they are also less likely to settle. That is, even after an opposition has been initiated, parties that have more to lose will expend significant time and money to win the case instead of compromising.

  11. Figures from this paragraph are based on our own calculations from the EUIPO raw data. Section 4 and Appendix A outline in detail the data compilation process.

  12. Scholars in the patent literature have examined extensively the determinants of patent pendency (time from application to grant) recognizing that uncertainty can play an important role in the marketplace (Palangkaraya et al. 2008).

  13. In the same spirit, early studies by von Graevenitz (2009) and Collette (2012) examine the strategic behavior of plaintiffs over potentially ‘harmful’ trademarks.

  14. Their interest was to examine whether distance accounts for the worldwide pattern of trademark flows and whether the latter is associated with trade and FDI. Their findings supported the strong role of distance in trademark flows and the latter as an integral part of the global economy.

  15. If this paper focused only on oppositions, then the origin and destination countries could be renamed to defendant and plaintiff countries. If it focused only on citations, then the origin and destination countries could be renamed to cited and citing countries.

  16. For both NUTS-3 regions and counties we also estimate gravity-like models with distance dummies. Results on the home bias for NUTS-3 and county are similar to the ones presented here and available upon request.

  17. https://euipo.europa.eu/ohimportal/en/open-data.

  18. Since this variable is no longer an integer, for econometric purposes it is rounded it to the nearest integer as its rest of properties (skewness, dispersion) warrant a Negative Binomial estimator.

  19. https://www.census.gov/geo/maps-data/data/tiger.html.

  20. Technology field 11 (Analysis of biological materials) is dropped due to zero observations. Hence, instead of 35, 34 technology fields are considered.

  21. A point that is useful to highlight is the insignificance of MarkProximity for opposition flows. This is likely to stem from the fact that MarkProximity from USPTO data has very little variation, especially from year to year. This is evident when comparing this metric between Tables 6 and 8 for EUIPO and USPTO data, respectively. Therefore, when we exclude year fixed effects and ΣTrademarkso and ΣTrademarksd the MarkProximity coefficient turns significant. Results are available upon request.

  22. Note that the estimation of Column 3 does not converge with all the county fixed effects. It therefore includes origin and destination US State fixed effects. A comparison of all the results of Table 5 with origin and destination US State fixed effects or with no fixed effects display qualitatively similar results and available upon request.

  23. http://www.wipo.int/classifications/nice/en/.

  24. https://stats.oecd.org/Index.aspx?DataSetCode=BTDIXE.

  25. These are the following. D01T03: Agriculture, forestry and fishing. D05T08: Mining of coal and lignite, extraction of crude petroleum and natural gas, mining of metal ores, other mining and quarrying. D10T12: Food products, beverages and tobacco. D13T15: Textiles wearing apparel, leather and related products. D16: Wood and products of wood and cork, except furniture. D17T18: Paper and printing. D19T22: Chemicals, rubber, plastics and fuel products. D23: Other non-metallic mineral products. D24T25: Basic metals and fabricated metal products, except machinery and equipment. D26T28: Machinery and equipment. D29T30: Transport equipment. D31T32: Furniture, other manufacturing. D36T99: Other activities. We drop D35 (Electricity, gas and water supply) due to insufficient data.

  26. http://www.geonames.org/.

  27. These twenty countries account for 71% of the trademark applications filed by the CNT38 group. There are nineteen countries from EU and CH. Specifically, these countries are: AT, BE, CH, CZ, DE, DK, ES, FI, FR, HR, HU, IT, LU, NL, PL, PT, SE, SI, SK and UK.

  28. https://ec.europa.eu/eurostat/web/gisco/geodata/reference-data/administrative-units-statistical-units/nuts.

  29. From the outset certain NUTS-3 regions were excluded that while administratively they may belong to the focal countries, geographically they are not close to Europe. For instance, French Guiana has its own NUTS-3 region but is located in South America. The countries with such NUTS-3 regions were FR, ES and PT.

  30. https://bulkdata.uspto.gov/data/trademark/dailyxml/ttab/.

References

  • Agrawal A, Cockburn I, Galasso A, Oettl A (2014) Why are some regions more innovative than others? The role of small firms in the presence of large labs. J Urban Econ 81:149–165

    Article  Google Scholar 

  • Allison PD, Waterman RP (2002) Fixed-effects negative binomial regression models. Soc Methodol 32(1):247–265

    Article  Google Scholar 

  • Andersson M, Ejermo O (2005) How does accessibility to knowledge sources affect the innovativeness of corporations?—evidence from Sweden. Ann Reg Sci 39(4):741–765

    Article  Google Scholar 

  • Antonelli C, Patrucco PP, Quatraro F (2011) Productivity growth and pecuniary knowledge externalities: an empirical analysis of agglomeration economies in European regions. Econ Geography 87(1):23–50

    Article  Google Scholar 

  • Arrow KJ (1962) The economic implications of learning by doing. Rev Econ Stud 29(3):155–173

    Article  Google Scholar 

  • Bacchiocchi E, Montobbio F (2010) International knowledge diffusion and home-bias effect: Do USPTO and EPO patent citations tell the same story? Scand J Econ 112(3):441–470

    Google Scholar 

  • Balland PA, Belso-Martínez JA, Morrison A (2016) The dynamics of technical and business knowledge networks in industrial clusters: Embeddedness, status, or proximity? Econ Geography 92(1):35–60

    Article  Google Scholar 

  • Belenzon S, Schankerman M (2013) Spreading the word: Geography, policy, and knowledge spillovers. Rev Econ Stat 95(3):884–903

    Article  Google Scholar 

  • Bertrand M, Duflo E, Mullainathan S (2004) How much should we trust differences-in-differences estimates? Q J Econ 119(1):249–275

    Article  Google Scholar 

  • Bloom N, Schankerman M, Van Reenen J (2013) Identifying technology spillovers and product market rivalry. Econometrica 81(4):1347–1393

    Article  Google Scholar 

  • Breschi S, Lissoni F (2009) Mobility of skilled workers and co-invention networks: an anatomy of localized knowledge flows. J Econ Geogr 9(4):439–468

    Article  Google Scholar 

  • Capello R, Caragliu A (2018) Proximities and the intensity of scientific relations: synergies and nonlinearities. Int Reg Sci Rev 41(1):7–44

    Article  Google Scholar 

  • Caragliu A, de Dominicis L, de Groot HL (2016) Both Marshall and Jacobs were right! Econ Geogr 92(1):87–111

    Article  Google Scholar 

  • Castaldi C (2019) All the great things you can do with trademark data: Taking stock and looking ahead. Strategic Organization

  • Castaldi C, Dosso M (2018) From R&D to market: using trademarks to capture the market capability of top R&D investors. Joint Research Centre (Seville site).

  • Castaldi C, Giarratana MS (2018) Diversification, Branding, and Performance of Professional Service Firms. J Service Res

  • Chinitz B (1961) Contrasts in agglomeration: New york and pittsburgh. Am Econ Rev 279–289

  • Crass D, Czarnitzki D, Toole AA (2019) The dynamic relationship between investments in brand equity and firm profitability: Evidence using trademark registrations. Int J Econ Business 26(1):157–176

    Article  Google Scholar 

  • Collette E (2012) Increasing Rival Costs: Trade-mark Opposition as a Tool for Entry-Deterrence. Carleton University, PhD diss.

    Google Scholar 

  • De Chernatony L, Dall’Olmo Riley F (1998) Defining a" brand": Beyond the literature with experts’ interpretations. J Market Manag 14(5):417–443

    Article  Google Scholar 

  • De Groot HL, Poot J, Smit MJ (2016) Which agglomeration externalities matter most and why? J Econ Surv 30(4):756–782

    Article  Google Scholar 

  • Drivas K, Economidou C, Karkalakos S, Tsionas EG (2016) Mobility of knowledge and local innovation activity. Eur Econ Rev 85:39–61

    Article  Google Scholar 

  • Drucker J, Feser E (2012) Regional industrial structure and agglomeration economies: an analysis of productivity in three manufacturing industries. Reg Sci Urban Econ 42(1–2):1–14

    Article  Google Scholar 

  • Eriksson RH (2011) Localized spillovers and knowledge flows: How does proximity influence the performance of plants? Econ Geogr 87(2):127–152

    Article  Google Scholar 

  • Feldman MP, Audretsch DB (1999) Innovation in cities: Science-based diversity, specialization and localized competition. Eur Econ Rev 43(2):409–429

    Article  Google Scholar 

  • Fischer MM, Griffith DA (2008) Modeling spatial autocorrelation in spatial interaction data: an application to patent citation data in the European Union. J Reg Sci 48(5):969–989

    Article  Google Scholar 

  • Fitjar RD, Rodríguez-Pose A (2013) Firm collaboration and modes of innovation in Norway. Res Policy 42(1):128–138

    Article  Google Scholar 

  • Fink C, Helmers C, Ponce CJ (2018a) Trademark squatters: Theory and evidence from Chile. Int J Ind Organ 59:340–371

    Article  Google Scholar 

  • Fink C, Fosfuri A, Helmers C, Myers AF (2018b) Submarine trademarks. Economic Research Working Paper No. 51. World Intellectual Property Organization (WIPO).

  • Glaeser EL, Kallal HD, Scheinkman JA, Shleifer A (1992) Growth in cities. J Polit Econ 100(6):1126–1152

    Article  Google Scholar 

  • Gourieroux C, Monfort A, Trognon A (1984) Pseudo maximum likelihood methods: Theory. Econ J Econ Soc, 681–700.

  • Greene, W. H., 1994. Accounting for excess zeros and sample selection in poisson and negative binomial regression models. NYU Working Paper No. EC-94–10.

  • Graham SJH, Hancock G, Marco AC, Myers AF (2013) The USPTO trademark case files dataset: descriptions, lessons, and insights. J Econ Manag Strat 22(4):669–705

    Article  Google Scholar 

  • Hidalgo, C.A. Klinger, B. Barabási, A.-L. and Hausmann, R. (2007). "The Product Space Conditions the Development of Nations", Science 317.

  • Jacobs J (1969) The Economy of Cities. Random House, New York

    Google Scholar 

  • Jaffe A (1986) Technological opportunity and spillovers of R&D: evidence from firms’ patents, profits, and market value. Am Econ Rev 76(5):984–1001

    Google Scholar 

  • Jaffe A, Trajtenberg M, Henderson R (1993) Geographic Localization of Knowledge Spillovers as Evidenced by Patent Citations. Q J Econ 108(3):577–598

    Article  Google Scholar 

  • Keller KL (1993) Conceptualizing, measuring, and managing customer-based brand equity. J Market 57(1):1–22

    Article  Google Scholar 

  • Kilmer PF (2013) The Value Equation of Trademark Oppositions: A Multinational Comparison of Costs and Perceived Benefits. INTA Bulletin, 68(5)

  • Kolympiris C, Kalaitzandonakes N, Miller D (2011) Spatial collocation and venture capital in the US biotechnology industry. Res Policy 40(9):1188–1199

    Article  Google Scholar 

  • Koval MR (2018) How Shorebilly Brewing Company Won the Trademark Battle, but Lost the War: A Cautionary Tale for Entrepreneurs. J Legal Stud Educ 35(1):45–82

    Article  Google Scholar 

  • Krugman, P. R. (1993). Geography and trade. MIT press.

  • Lambert D (1992) Zero-inflated poisson regression, with an application to defects in manufacturing. Technometrics 34(1):1–14

    Article  Google Scholar 

  • Lucking B, Bloom N, Van Reenen J (2018) Have R&D Spillovers Changed? (No. w24622). National Bureau of Economic Research.

  • Lychagin S, Pinkse J, Slade ME, Reenen JV (2016) Spillovers in space: Does geography matter? J Ind Econ 64(2):295–335

    Article  Google Scholar 

  • Malecki EJ (2010) Everywhere? The geography of knowledge. J Reg Sci 50(1):493–513

    Article  Google Scholar 

  • Malmberg A, Maskell P (2002) The elusive concept of localization economies: towards a knowledge-based theory of spatial clustering. Environ Planni A Econ Space 34(3):429–449

    Article  Google Scholar 

  • Maraut S, Dernis H, Webb C, Spiezia V, Guellec D (2008) The OECD REGPAT database: a presentation

  • Marshall A (1890) Principles of economics. Macmillan, London

    Google Scholar 

  • Mayer, Thierry, and Soledad Zignago (2011) Notes on CEPII’s distances measures: The GeoDist database

  • Mazzola F, Cascio IL, Epifanio R, Di Giacomo G (2018) Territorial capital and growth over the Great Recession: a local analysis for Italy. Ann Reg Sci 60(2):411–441

    Article  Google Scholar 

  • Mendonça S, Pereira TS, Godinho MM (2004) Trademarks as an indicator of innovation and industrial change. Res Policy 33(9):1385–1404

    Article  Google Scholar 

  • Mendonça S (2014) National adaptive advantages: Soft innovation and marketing capabilities in periods of crisis and change. In: Structural Change, Competitiveness and Industrial Policy (pp. 149–166). Routledge

  • Morescalchi A, Pammolli F, Penner O, Petersen AM, Riccaboni M (2015) The evolution of networks of innovators within and across borders: evidence from patent data. Res Policy 44(3):651–668

    Article  Google Scholar 

  • Morrin M, Jacoby J (2000) Trademark dilution: empirical measures for an elusive concept. J Public Policy Market 19(Fall):265–276

    Article  Google Scholar 

  • Mowery DC, Ziedonis AA (2015) Markets versus spillovers in outflows of university research. Res Policy 44(1):50–66

    Article  Google Scholar 

  • Palangkaraya A, Jensen PH, Webster E (2008) Applicant behaviour in patent examination request lags. Econ Lett 101(3):243–245

    Article  Google Scholar 

  • Panzera D, Postiglione P (2014) Economic growth in Italian NUTS 3 provinces. Ann Reg Sci 53(1):273–293

    Article  Google Scholar 

  • Peri G (2005) Determinants of Knowledge Flows and Their Effect on Innovation. Rev Econ Stat 87(2):308–322

    Article  Google Scholar 

  • Plummer LA, Acs ZJ (2014) Localized competition in the knowledge spillover theory of entrepreneurship. J Bus Ventur 29(1):121–136

    Article  Google Scholar 

  • Porter M (1990) The competitive advantage of nations. Macmillan, London

    Book  Google Scholar 

  • Rodríguez-Pose A (1999) Innovation prone and innovation averse societies: economic performance in Europe. Growth Change 30(1):75–105

    Article  Google Scholar 

  • Romer PM (1986) Increasing returns and long-run growth. J Polit Econ 94(5):1002–1037

    Article  Google Scholar 

  • Rosenthal SS, Strange WC (2001) The determinants of agglomeration. J Urban Econ 50(2):191–229

    Article  Google Scholar 

  • Schautschick P, Greenhalgh C (2016) Empirical studies of trade marks–the existing economic literature. Econ Innov New Technol 25(4):358–390

    Article  Google Scholar 

  • Schmoch U (2003) Service marks as novel innovation indicator. Res Eval 12(2):149–156

    Article  Google Scholar 

  • Schmoch U (2008) Concept of a technology classification for country comparisons. Final report to the world intellectual property organisation (wipo), WIPO

  • von Graevenitz G (2009) Which reputations does a brand owner need? Evidence from Trade Mark Opposition

  • Zolas N, Lybbert TJ, Bhattacharyya P (2017) An ‘Algorithmic Links with Probabilities’ Concordance for Trademarks with an Application Towards Bilateral IP Flows. World Econ 40(6):1184–1213

    Article  Google Scholar 

Download references

Acknowledgements

I would like to thank Fragiskos Archontakis, Dominic Berry, Berris Charnley, Christos Kolympiris, Zhen Lei, Alfons Palangkaraya, Andreas Panagopoulos and Nathan Wajsman; also, conference participants of the European Progress Conference 2018. All errors are my own.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Kyriakos Drivas.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendices

Appendix A: data details

1.1 Compilation of the product-related proximity metrics

Starting with the MarkProximity, the NICE classifications for every trademark are collected. When applicants file for trademarks, they also claim the NICE classifications their trademark can be used. The NICE classification scheme categorizes the entire business spectrum in 45 distinct classes.Footnote 23 For an applicant to register the trademark with a particular NICE classification, she needs to show that she is using it commercially in the specific class. Therefore, these NICE classes are associated with actual business activity in each of these classes. First, the frequency of trademark applications for each class for each year for each country is calculated. In case the trademark application claims more than one NICE class, is double counted. Then, the share each NICE class has in the country’s overall trademark portfolio in each year is computed. Thus, the vector of shares for a country p is denoted as shtmp and each element in the vector as shtmp,nwhere n denotes the nth class. The MarkProximity for countries p and d per year is then calculated as:

$$Markproximity = \frac{{shtm_{p}^{'} *shtm_{d} }}{{\sqrt {\mathop \sum \nolimits_{{n = 1}}^{{45}} shtm_{{n,s}}^{2} \mathop \sum \nolimits_{{n = 1}}^{{45}} shtm_{{d,n}}^{2} } }}$$

Note that we construct the same index for the pairs of US States based on USPTO’s trademark data.

For the country-level case, we also construct an export similarity index motivated by the work of Hidalgo et al. (2007). To construct the export proximity measure, a similar approach is used. From the OECD STAN databaseFootnote 24 data are downloaded, for every country in the sample, on their exports by industry sector, as defined by the International Standard Industrial Classification. To minimize missing observations, the industry sectors are bundled in thirteen industry sectors per OECD’s guidelines – hereafter these bundled sectors are referred to as the industry sectors.Footnote 25 For each of these industry sectors their export share per country per year is calculated. Thus the vector of shares for a country p is denoted as shexp and each element in the vector as shexp,n where n denotes the nth industry sector. The ExportProximity per year for countries p and d is then calculated as:

$$ExportProximity = \frac{{shex_{p}^{'} *shex_{d} }}{{\sqrt {\mathop \sum \nolimits_{{n = 1}}^{{13}} shex_{{p,n}}^{2} \mathop \sum \nolimits_{{n = 1}}^{{13}} shex_{{d,s}}^{2} } }}$$

As with the case of MarkProximity, ExportProximity is the uncentered correlation of export profiles between two countries.

1.2 EUIPO data compilation process

The data were downloaded during 2018 effectively acquiring all trademark applications filed during 1996–2016. After dropping a few hundred erroneous observations, 1,348,195 trademark applications were obtained. In the initial population of the 1,348,195 trademark applications, 209,164 were opposed at least once until the end of 2016. Thus, the overall probability of opposition is 15.1%. This can roughly be translated as: one every seven trademark applications in EUIPO was opposed at least once.

Of this initial population of trademark applications, 12% of them lack information on the applicant (either the name, identifier or country of origin is missing). Therefore, the sample reduces to 1,217,886. For this subset, the number of opposed trademark applications is equal to 183,173 which amounts to 183,173/1,217,886 = 15.04% a percentage very similar to the above. Further, 21,583 opposed trademark applications are excluded; for these applications, there is no sufficient information for the plaintiff, or at least one plaintiff in case multiple oppositions have occurred. Therefore, the sample of opposed trademark applications is reduced to 161,590.

Of the 1,217,886 trademark applications, the focal applications that are kept belong to an applicant that is located in one the EU28 or one of the ten largest applicants, excluding EU28; these 38 countries are denoted as the CNT38 group. After this data cut, the sample is reduced to 1,162,800; a simple division of the above two figures shows that these applicants account for 95% of all trademark applications, with non-missing owner, at the EUIPO during 1996–2016. Also note, that this latter sample is employed to construct the ΣTrademarksp,t−5,t and ΣTrademarksp,t−5,t indicators for Eq. 1.

The CNT38 group is also responsible for the overwhelming majority of oppositions. As defendant, it shows up in 94.5% of the oppositions while as a plaintiff in 97.5%. Overall, of the 161,590 opposed trademarks with full information, 149,207 disclose defendant and plaintiff from the CNT38 group, i.e., 92%. This sample of 149,207 trademark applications is the baseline which is used for the country-level analysis.

1.3 NUTS-3 data compilation process

From the EUIPO trademark data, an additional crucial piece of information was collected; the zip code for the trademark applicant and the plaintiff. This additional piece of information is collected only for the EU28 countries and Switzerland focusing, therefore, on Europe.

The second task was to find the centroid of each zip code. Data were downloaded from the GeoNames project.Footnote 26 GeoNames collects and maintains a multitude of locations information on zip codes and their centroids. Next, for each of the above 29 countries we extracted their individual zip code files and matched them with the zip codes from the trademark data. We only kept countries where more than 85% of applicants and plaintiffs could be matched with a zip code in the GeoNames database. To this end, 20 countries that satisfied that threshold were kept.Footnote 27 These countries are denoted as REGION20.

For the third step, we downloaded from Eurostat the shapefiles for Europe’s NUTS-3 regionsFootnote 28 opting for the 2010 classification.Footnote 29 Fourth, each applicant and plaintiff are assigned to one of the above NUTS-3 regions based on their zip code’s centroid. Overall, of the total 98,129 trademark applications where the plaintiff and defendant are from the REGION20 countries, we obtained accurate NUTS-3 location information for both the plaintiff and defendant for 90,084 trademark applications.

1.4 USPTO data compilation process

The USPTO has released every case of the Trademark Trial and Appeal Board, the division in charge of handling opposition cases, in a single xml file.Footnote 30 After an elaborate data cleaning, any relevant information was collected. The second dataset, with all the necessary trademark bibliographic information comes again from the USPTO and the Office of the Chief Economist (Graham et al. 2013). During the period 1996–2016, there are 4,867,600 trademark applications that have been filed by the US located entities. This constitutes the 83% of trademark applications that were filed at the USPTO during the same period.

During the same time period, as with the EUIPO data (1996–2016), 105,826 trademark applications were opposed. As the USPTO data disclose correspondents’ location information one can readily extract the State for both the defendant and plaintiff. With this this in mind, 100,265 trademark applications that have been opposed between entities located in the 48 contiguous US states (plus DC) are identified.

Appendix B: Supplementary tables and figures

See Table

Table 6 Summary statistics at the country level

6,

Table 7 Summary statistics at the NUTS-3 level – EUIPO and EPO data

7,

Table 8 Summary statistics at the US State level – USPTO data

8,

Table 9 Summary statistics at the US County level—USPTO data

9,

Table 10 Re-estimate Table 5 via Poisson regressions

10,

Table 11 Re-estimate Table 1—Columns 1–4. Replace Oppositions with OppositionsWeighted

11,

Table 12 Counterpart to Table 1—Estimate it via zero-inflated negative binomial model

12.

Appendix C: EU accession

There were three waves of EU accession within the studied time period that merit attention: 2004, 2007 and 2013. In 2004, ten countries entered EU; namely CY, CZ, EE, HU, LV, LT, MT, PL, SK and SI. In 2007 BG and RO entered while in 2013 HR entered. With this in mind, Eq. 1 can be re-formulated as follows:

$$Y_{{p,d,t}} = \beta _{0} + \beta _{1} OwnCountry_{{o,d}} + \beta _{2} Year_{{2004}} + \beta _{3} OEntrant2004_{o} *Year_{{2004}} + \beta _{4} DEntrant2004_{d} *Year_{{2004}} + \beta _{5} Year_{{2007}} + \beta _{6} OEntrant2007_{o} *Year_{{2007}} + \beta _{7} DEntrant2007_{d} *Year_{{2007}} + \beta _{8} Year_{{2013}} + \beta _{9} OEntrant2013_{o} *Year_{{2013}} + \beta _{{10}} DEntrant2013_{d} *Year_{{2013}} +$$

\(Origin_{o} + Destination_{d} + \beta _{{11}} \sum IP_{{o,t - 5,t}} + \beta _{{12}} \sum IP_{{d,t - 5,t}} + \beta _{{13}} ExportProximity_{{o,d,t}} + \beta _{{13}} MarkProximity_{{o,d,t}} + ~\varepsilon _{{p,d,t}}\)Year2004 takes the value of 1 for years after 2004 and 0 otherwise; similarly for Year2007 and Year2013. OEntrant2004o takes the value of 1 if the origin country is one of the countries that entered in 2004 and 0 otherwise. DEntrant2004d takes the value of 1 if the destination country is one of the countries that entered in 2007 and 0 otherwise. A similar definition applies to OEntrant2007o, DEntrant2007d, OEntrant2013o and DEntrant2013d.

The focus of this specification is on the interaction terms. For instance, β3 shows to what extent countries that entered in the EU in 2004 received more oppositions/citations after 2004 compared to before. Similarly, β4 shows how countries that entered in the EU in 2004 altered their opposition/citation behavior as plaintiffs/citers after 2004 compared to before.

Results are displayed in Table 13 Columns 1 considers oppositions while Column 2 patent citations. The interaction terms for the opposition case are positive and significant at the 1%. Further, they are roughly similar in size. This indicates that upon accession these countries increased their opposition rates both as plaintiffs and defendants. Results for citations are distinctively different. Focusing on the first wave, entrants cited more but were not cited. As for the rest two waves the picture is not as clear. This could be due to the fact that follow-on innovation can take a few years to materialize and therefore observed as patent citations.

Table 13 Examine the three waves of EU Accession to the new entrants

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Drivas, K. Which travels farther? Knowledge or rivalry?. Ann Reg Sci 67, 299–333 (2021). https://doi.org/10.1007/s00168-021-01049-y

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00168-021-01049-y

JEL Classification

Navigation