Nagaoka Nishimura (2006) - An Empirical Assessment Of The Effects Of Patent Thickets

Article
Has bibtex key
Has article title	An Empirical Assessment Of The Effects Of Patent Thickets
Has author	Nagaoka Nishimura
Has year	2006
In journal
In volume
In number
Has pages
Has publisher
© edegan.com, 2016

• This page is referenced in the Patent Thicket Literature Review
• This page is listed on the PTLR Core Papers page

Reference

• Nagaoka, S. and Nishimura, Y. (2006), "An Empirical Assessment of the Effects of Patent Thickets", unpublished, Hitotsubashi University

@article{nagaoka2006empirical,
  title={An Empirical Assessment of the Effects of Patent Thickets},
  author={Nagaoka, S. and Nishimura, Y.},
  journal={unpublished, Hitotsubashi University},
  year={2006},
  abstract={This paper investigates the effects of patent thicket, exploiting a unique measure of the patent thicket from an inventor survey and on data on the use of patents by around 1,200 Japanese firms. First, we develop an index of patent thicket in each industry, mainly using the size of complementary patents necessary for commercializing an invention, which is from the RIETI inventor survey. Secondly, we will assess empirically how “patent thickets” affect the patenting propensity and the use of the patents as well as whether it negatively affects the appropriability of the R&D. In such assessment, we focus on the interaction effects between the level of patent thickets and firm size, based on a theoretical view which patent thickets affect a large firm more, since the latter has a large stake for the prevention of a hold-up and has a larger demand for the combining patents. Focusing on the firm-industry interaction effects allow us to control for industry fixed effects. Major findings are as follows. Firstly, we have found that patent thickets are high in information and communication electronics equipment sector, information and communication sector and motor vehicles sectors while they are low in printing and food products industries. These results based on the average number of patents joint used for commercializing an invention are largely consistent with existing empirical studies on patent thickets. Secondly, we have found that patenting propensity increases more with firm size in the industry where patent thicket is high, controlling for industry fixed effects. Such effects do not exist among small firms. On the other hand, our findings suggest that a number of invention disclosures to the IPR department or that for applying a patent by a firm do not increase significantly more with firm size in the industry where patent thicket is high. Thirdly, we have found that while the proportion of used patents declines with firm † We would like to thank for valuable comments by the members of the research project at the Institute of Intellectual Property. We would like to thank for the Japan Patent Office for enabling us to use the firm level data of the Survey of Intellectual Property-Related Activities. 2 size, such effect is no stronger in the industry where patent thicket is high. One explanation for this is that cross-licensing is often a response for patent thicket problem. Cross licensing facilitates access to complementary technologies of other firms, which makes commercialization of own technology easier. We also find that while blocking increases with firm size, such effect is no stronger in the industry where patent thicket is high. We do not observe significantly larger negative effects of patent thickets on the patent use by every size of firms. Lastly, we have found that exclusively used patents make the largest contributions to a firms’ profit, in particular, more than cross licensed patents.},
  discipline={Mgmt},
  research_type={Empirical},
  industry={},
  thicket_stance={},
  thicket_stance_extract={},
  thicket_def={},
  thicket_def_extract={},  
  tags={},
  filename={Nagaoka Nishimura (2006) - An Empirical Assessment Of The Effects Of Patent Thickets.pdf}
}

File(s)

• Download the PDF
• Repository record

Abstract

This paper investigates the effects of patent thicket, exploiting a unique measure of the patent thicket from an inventor survey and on data on the use of patents by around 1,200 Japanese firms. First, we develop an index of patent thicket in each industry, mainly using the size of complementary patents necessary for commercializing an invention, which is from the RIETI inventor survey. Secondly, we will assess empirically how “patent thickets” affect the patenting propensity and the use of the patents as well as whether it negatively affects the appropriability of the R&D. In such assessment, we focus on the interaction effects between the level of patent thickets and firm size, based on a theoretical view which patent thickets affect a large firm more, since the latter has a large stake for the prevention of a hold-up and has a larger demand for the combining patents. Focusing on the firm-industry interaction effects allow us to control for industry fixed effects. Major findings are as follows. Firstly, we have found that patent thickets are high in information and communication electronics equipment sector, information and communication sector and motor vehicles sectors while they are low in printing and food products industries. These results based on the average number of patents joint used for commercializing an invention are largely consistent with existing empirical studies on patent thickets. Secondly, we have found that patenting propensity increases more with firm size in the industry where patent thicket is high, controlling for industry fixed effects. Such effects do not exist among small firms. On the other hand, our findings suggest that a number of invention disclosures to the IPR department or that for applying a patent by a firm do not increase significantly more with firm size in the industry where patent thicket is high. Thirdly, we have found that while the proportion of used patents declines with firm † We would like to thank for valuable comments by the members of the research project at the Institute of Intellectual Property. We would like to thank for the Japan Patent Office for enabling us to use the firm level data of the Survey of Intellectual Property-Related Activities. 2 size, such effect is no stronger in the industry where patent thicket is high. One explanation for this is that cross-licensing is often a response for patent thicket problem. Cross licensing facilitates access to complementary technologies of other firms, which makes commercialization of own technology easier. We also find that while blocking increases with firm size, such effect is no stronger in the industry where patent thicket is high. We do not observe significantly larger negative effects of patent thickets on the patent use by every size of firms. Lastly, we have found that exclusively used patents make the largest contributions to a firms’ profit, in particular, more than cross licensed patents.

Review

Measures of thicket

The paper develops a patent thicket index for 24 industrial sectors covering around 1,200 Japanese firms.

• The index is based on the average size of complementary patents necessary for commercializing an invention/product (in each sector)
  • The "size" measure is collected from the RIETI (Research Institute of Economics, Trade and Industry) Inventor Survey.
  • While not explained in the article, other articles on the RIETI website indicate that "size" means the bundle of all patents necessary to commericalize a firm's product/invention, whether those patents are owned by the firm or licensed from other firms.
• Following Ziedonis, a fragmentation index (1-HHI) of patent stock ownership in each industrial sector is used to weight the average size of complementary patents in each sector.
  • Patent stock ownership is obtained from SIPRA (Survey of Intellectual Property-Related Activites) conducted by the JPO (Japanese Patent Office) for about 1,200 major Japanese firms n the manufacturing, and information and communications sector from 2007 to 2010.
• The index ranges from 2.5 for the Printing sector to 48.4 for the Infomation and communication electronics equipment sector.

Sample

• About 6,000 observations on Japanese manufacturing firms from 2007-2010 in 25 industrial sectors.

Results

Distribution of Patent Thickets:

patent thickets are high in information and communication sector, information and communication electronics equipment sector and motor vehicles sectors while they are low in printing and construction industries."

Patenting Propensity:

"...a larger firm tends to have a higher patenting propensity due to a patent thicket since a large firm is more likely to suffer from hold-up problem since it makes large sunk investments."

• The firm's stock of patents is significant positively related to the patent thicket index interacted with number of employees in firm;
• No impact is found for smaller firms.

Patent Use:

"...firms avoid patenting in order to prevent other firms for accessing their inventions through cross-licensing."

• Interaction terms between the patent thicket index and number of inventions held in secrecy, and number of inventions acquired are both significantly positively related to patent thicket measures.
• The number of invention disclosures and patent applications is unrelated to firm size.

"patent thicket does not have any significant impact on the patent use and in an industry with high thicket, and Japanese firms can easily access complementary inventions of other firms' through cross-licensing and blocking becomes less effective in an industry with higher patent thicket."

• The interaction between patent thicket index and number of employees (as wells as R&D per employee) are significantly positively related to cross-licensing, but significantly negatively related to blocking.

Social Welfare Consequences

Patent Thicket impact on profitability:

"...no strong evidence that R&D spending has signficantly smaller impacts on a firm's profitability in the industry in which a patent thicket is important...patent thicket only affect[s] the composition of patent exploitation such as the frequency of cross-licensing and blocking and these results suggest that this change in the composition of patent exploitation can have either positive or negative influence on firm’s profitability to some extent in Japan.

• Exclusive patent use is signfiicantly positively related to profits, with a larger coefficient than either cross-licensing, blocking or open use of patents (coefficients of 2.4, 1.2, 2.0. 2.4).
• However, interactions of different modes of use and the patent thicket index are negative, but relatively small (coefficients of -.12 to -.016).

Dependent Variables and Model

A number of dependent variables are analyzed using negative binomial regressions with clustering by firm:

• Firms' patent stocks (number of patents held);
• The number of inventions by the firm;
• The number of inventions by the firm that are kept in secrecy;
• The number inventions for which there are patent applications.

A number of dependent variables are analyzed using pooled OLS with clustering by firm:

• Rate of internal use of own patents;
• Rate of exclusive internal use of own patents;
• Rate of cross-licensed patents;
• Rate of blocking patents;
• Rate of unused patents.

Profitability as measured by operating income per employee across the 2008-2010 period is analyzed using pooled OLS.

• The patent thicket measures used as explanatory variables are:
  • An interaction between the patent thicket index and number of employees in the firm, because it is hypothesized that larger firms hold larger patent portfolios;
  • An interaction between the patent thicket index and R&D expenditures per employee.
  • For the profitability analyses, a series of interactions between the above patent thicket measures and modes of patent exploitaton(used, unused, exclusive use, share of portfolio cross-licensed, rate used to block, open use)

• Models also control for:
  • number of firm employees;
  • R&D expenditures per employee;
  • sales revenue per employee;
  • industry and year indicators, as well as interaction terms.