Cockburn MacGarvie Muller (2010) - Patent Thickets Licensing And Innovative Performance

From edegan.com
Jump to navigation Jump to search
Article
Has bibtex key
Has article title Patent Thickets Licensing And Innovative Performance
Has author Cockburn MacGarvie Muller
Has year 2010
In journal
In volume
In number
Has pages
Has publisher
© edegan.com, 2016

Reference

  • Cockburn, I.M., MacGarvie, M.J. and Müller , E. (2010), "Patent thickets, licensing and innovative performance", Industrial and Corporate Change, Vol.19, No.3, pp.899--925
@article{cockburn2010patent,
  title={Patent thickets, licensing and innovative performance},
  author={Cockburn, I.M. and MacGarvie, M.J. and M{\"u}ller, E.},
  journal={Industrial and Corporate Change},
  volume={19},
  number={3},
  pages={899--925},
  year={2010},
  abstract={We examine the relationship between fragmented intellectual property (IP) rights and the innovative performance of firms, taking into consideration the role played by in-licensing of IP. We find that firms facing more fragmented IP landscapes have a higher probability of in-licensing. We observe a negative relationship between IP fragmentation and innovative performance, but only for firms that engage in in-licensing. In contrast, greater IP fragmentation is associated with higher innovative performance for firms that do not in-license. Furthermore, the effects of fragmentation on innovation also appear to depend on the size of a firm’s patent portfolio. These results suggest that the effects of fragmentation of upstream IP rights are not uniform, and instead vary according to the characteristics of the downstream firm.},
  discipline={Mgmt, Econ},
  research_type={Empirical},
  industry={},
  thicket_stance={},
  thicket_stance_extract={},
  thicket_def={},
  thicket_def_extract={},  
  tags={},
  filename={Cockburn MacGarvie Muller (2010) - Patent Thickets Licensing And Innovative Performance.pdf}
}

File(s)

Abstract

We examine the relationship between fragmented intellectual property (IP) rights and the innovative performance of firms, taking into consideration the role played by in-licensing of IP. We find that firms facing more fragmented IP landscapes have a higher probability of in-licensing. We observe a negative relationship between IP fragmentation and innovative performance, but only for firms that engage in in-licensing. In contrast, greater IP fragmentation is associated with higher innovative performance for firms that do not in-license. Furthermore, the effects of fragmentation on innovation also appear to depend on the size of a firm’s patent portfolio. These results suggest that the effects of fragmentation of upstream IP rights are not uniform, and instead vary according to the characteristics of the downstream firm.

Review

Measures of thicket

Measures of patent thickets include:

  • A technology fragmentation index that follows the approach in Ziedonis (2004), but is defined at technology rather than firm level.
    • The index is based on technology shares defined by the number of references to other firm's patents in a given firm's patent portfolio divided by the total number of references in that firm's portfolio;
    • The index is calculated using patents filed in the past 3 years by larger firms from any countries (not just Germany) that are recorded by European Patent Office (EPO);
    • The index averages ranges from 0.607 to 0.794 across the 30 technology classes in the analysis.

Sample

Size, data source, industries, geography:

  • The paper analyzes a sample of about 1,100 (varyies depending on the analysis) small and medium-sized German manufacturing companies with at least one patent drawn the Mannheim Innovation Panel survey.
  • Imposing selection criteria, excluding outliers and observations with missing data results in typically one observation per firm and spans years 1993-2005.

Results

Licensing expenditures:

"Firms facing more fragmented IP landscapes have higher licensing costs, largely because they are more likely to in-license. In-licensing firms with small patent portfolios also have higher licensing expenditures (at the 10% level). This is broadly consistent with a “royalty stacking” story, and/or a relationship between bargaining problems and fragmented ownership of rights."
  • Fragmentation significantly increases licensing expenditures with a marginal effect of 1.051%, a result driven by firms that in-license and have fewer than 5 patents.


Innovative Performance (share of sales from new products):

"Overall, the estimated relationship between fragmentation and innovative performance is positive and marginally significant."
  • Fragmentation is significantly positively related to the share of sales new to the market, and is marginally significant for the share of sales new to the firm (Tobit coefficients of 78 and 50 respectively).
"[However, t]hese regressions reveal that the negative relationship between fragmentation and innovation is strongest for licensing firms with fewer patents. This result is intriguing, as it is consistent with the hypothesis that in-licensing firms with smaller patent portfolios are more susceptible to the type of hold-up associated with patent thickets, while firms with large portfolios that do not need to license upstream technology may actually benefit from the existence of patent thickets."'
  • This positive relationship between fragmentation and innovative performance is driven by non-licensees (whose Tobit coefficients are 116 and 78 respectively)
    • In contrast, the share of sales new to the firm for licensees shows an opposite negative effect (with Tobit coefficient of -120).
    • Further analysis of innovation measures indicates that the negative impact of fragmentation on innovation is associated with "outsider" firms (with less than 5 patents).

Social Welfare Consequences

"...it is striking that we do not observe any significant negative correlation between fragmentation nd innovative sales more generally, and we even estimate a positive correlation between fragmentation and innovative sales among firms that do not n-license. This heterogeneity in the estimated effects of fragmentation suggests that here is no straightforward, uniform relationship between patent thickets and innovation or all firms.

Dependent Variable and Model

  • Dependent Variables analyzed include:
    • Licensing expenditures as a percentage of sales is analyzed using a Tobit model (since most firms have no such expenditures) accounting for firm and industry characteristics;
    • The share of sales new to the market, and share of sales new to the firm are analyzed in order to assess the impact of patent thickets on innovation in the market and firm.
  • Models control for both dependent variables control for the following variables:
    • Patent stock of firm/employees;
    • Number of employees in firm;
    • Age of firm;
    • Indicators for each industry and year and whether firm is based in East Germany;
    • Herfindahl index of industry concentration in Germany;
    • Number of patents held by the firm at technology level;
    • The licensing expenditures model also controls for R&D as a percent of sales, while the probability of in-licencing model controls for innovation expenditures as a percent of sales (and its square).