Source-linked AI summary
Determinants of Patent Citations in Biotechnology: An Analysis of Patent Influence Across the Industrial and Organizational Boundaries
Antonio Messeni Petruzzelli, Daniele Rotolo, Vito Albino
TL;DR
The paper addresses limited knowledge about how patent-influence antecedents vary across industrial and organizational boundaries. It distinguishes patent influence across these boundaries and finds that the six determinants contribute differently across domains, including contrasting effects of scientific knowledge.
Problem
Knowledge remains limited on how patent-influence antecedents affect subsequent technologies across different industrial and organizational boundaries.
Method
The analysis distinguishes patent influence according to biotechnology and organizational boundaries, measuring whole patent influence by subsequent citations.
Results
The six determinants contribute differently across industrial and organizational domains; scientific knowledge reduces influence outside biotechnology but increases relevance for the assignee’s future inventions.
Takeaways & Limitations
Patent influence should be analyzed separately across industrial and organizational boundaries because determinant effects differ by domain.
Takeaways & Limitations
Differences across patent-impact dimensions may also reflect characteristics of the assignee.
Abstract
from arXiv · showhide
The present paper extends the literature investigating key drivers leading certain patents to exert a stronger influence on the subsequent technological developments (inventions) than other ones. We investigated six key determinants, as (i) the use of scientific knowledge, (ii) the breadth of the technological base, (iii) the existence of collaboration in patent development, (iv) the number of claims, (v) the scope, and (vi) the novelty, and how the effect of these determinants varies when patent influence - as measured by the number of forward citations the patent received - is distinguished as within and across the industrial and organizational boundaries. We conducted an empirical analysis on a sample of 5671 patents granted to 293 US biotechnology firms from 1976 to 2003. Results reveal that the contribution of the determinants to patent influence differs across the domains that are identified by the industrial and organizational boundaries. Findings, for example, show that the use of scientific knowledge negatively affects patent influence outside the biotechnology industry, while it positively contributes to make a patent more relevant for the assignee's subsequent technological developments. In addition, the broader the scope of a patent the higher the number of citations the patent receives from subsequent non-biotechnology patents. This relationship is inverted-U shaped when considering the influence of a patent on inventions granted to other organizations than the patent's assignee. Finally, the novelty of a patent is inverted-U related with the influence the patent exerts on the subsequent inventions granted across the industrial and organizational boundaries.
1. INTRODUCTION
The paper examines why some patents exert greater influence on subsequent technological developments and whether six determinants operate differently within and across industrial and organizational boundaries. Using 5,671 patents from 293 US biotechnology firms, it finds domain-specific effects.
- Patent influence is commonly assessed through forward citations to subsequent technological developments.
- The study addresses limited knowledge about whether determinants of patent influence vary across specific industrial and organizational domains.
- Six determinants are examined: scientific knowledge, technological-base breadth, collaboration, claims, patent scope, and novelty.
- The analysis covers 5,671 granted USPTO patents registered by 293 US biotechnology firms from 1976 to 2003.
- Scientific knowledge negatively affects influence outside biotechnology but positively contributes to the assignee’s future inventions.
- Broader patent scope increases citations from subsequent non-biotechnology patents, is inverted-U shaped for influence on other organizations, and novelty is inverted-U related across both boundaries.
2. THEORETICAL BACKGROUND
The theoretical background explains why forward citations are used to assess patent influence and emphasizes that influence should be studied across industrial and organizational boundaries. It reviews six potential determinants, including scientific knowledge, search breadth, collaboration, claims, scope, and novelty.
- Forward citations are widely used as indicators of patent influence and are associated with technical, economic, and strategic patent value.
- Patent citations indicate the technological base on which a patent builds, but examiner involvement can introduce noise when citations proxy knowledge flows.
- Forward citations remain among the most suitable indicators of patent influence despite this measurement limitation.
- Prior research has paid less attention to where patent influence occurs, including whether subsequent inventions remain within the industry or cross industrial and organizational boundaries.
- The paper focuses on six determinants: scientific knowledge, technological-base breadth, collaboration, claims, scope, and novelty.
- Evidence on scientific knowledge and patent impact is mixed, while broad technological search may involve risks for successful invention development.
3. METHODS
The study analyzes patent influence in the US biotechnology industry using forward citations and distinguishes influence across industrial and organizational boundaries. It examines six patent-level determinants in a sample of 5,671 patents from 293 firms.
- Research setting and measurement: The study addresses concerns about unobserved heterogeneity across industries by focusing on biotechnology, where innovations may be covered by relatively few patents.The authors argue that discrete product fragmentation and potentially substantial rents from commercialization or licensing support patent data as an R&D-outcome proxy in this setting.
- Dependent variables: Patent influence is measured by forward citations received from subsequent patents within five years of filing.The measure includes citations from following patents up to five years after the focal patent’s filing date.
- Dependent variables: Influence is separated across biotechnology and non-biotechnology patents and across citations from the same versus different assignees.Industrial classification uses nine US technology classes defining biotechnology; organizational influence distinguishes the same firm from other assignees.
- Robustness and measurement: Citation-based influence measures are strongly related to patent family size, supporting their use as indicators of invention relevance.Family size is defined as the number of patent systems in which protection for one invention is sought, although it cannot distinguish influence across domains.
- Independent variables: The analysis examines scientific knowledge, technological breadth, joint development, claims, scope, and novelty as determinants of patent influence.Scientific knowledge is measured using non-patent references, while technological breadth uses the distribution of cited patents across technological classes.
4. RESULTS
The results show that patent determinants vary across industrial and organizational domains of influence. Claims and novelty often have inverted-U relationships with influence, while scientific knowledge, technological breadth, collaboration, and scope differ by domain.
- Estimation procedure: Negative binomial estimation is preferred because the over-dispersion parameter is significantly different from zero in each model.The authors also remove statistically insignificant squared terms to reduce multicollinearity risk.
- Overall influence: Claims and novelty have inverted-U relationships with overall patent influence.In Model 1, Claims is positive and Claims2 negative, while Novelty is positive and Novelty2 negative.
- Industrial boundaries: Within biotechnology, joint development positively affects citations, while claims retain an inverted-U relationship with influence.More than one firm participating in patent development is positively associated with citations from subsequent biotechnology inventions.
- Industrial boundaries: Outside biotechnology, scientific knowledge and technological breadth negatively affect patent influence.Claims and scope have positive effects on citations from subsequent non-biotechnology patents.
- Industrial boundaries: Novelty has an inverted-U relationship with influence outside the biotechnology industry.The reported novelty coefficient is positive and the squared novelty coefficient is negative in Model 3.
- Organizational boundaries: Within the assignee’s organization, scientific knowledge is positively related to patent influence and claims have an inverted-U effect.The scientific-knowledge coefficient is positive, while the claims coefficient is positive and its squared term is negative.
- Organizational boundaries: Across organizational boundaries, claims, scope, and novelty exert non-monotonic effects on citations from other firms.Claims and novelty show positive linear terms with negative squared terms; scope follows the same inverted-U pattern.
- Control variables: Team size is generally positively associated with patent impact, whereas the age variation of cited patents is generally negatively associated with it.These control-variable effects do not apply to influence across the industrial field or within the organization.
5. DISCUSSION
The six determinants contribute differently to patent influence across industrial and organizational domains. Scientific knowledge, technological breadth, collaboration, claims, scope, and novelty show domain-specific relationships with subsequent patent citations.
- Scientific knowledge: Scientific knowledge negatively affects citations from non-biotechnology patents but increases relevance for the assignee’s subsequent inventions.A broader scientific knowledge base may reduce understanding and applicability outside biotechnology while providing assignees with more knowledge resources.
- Technological breadth: Technological breadth generally has no significant effect, but greater breadth negatively affects influence outside biotechnology.Complexity may outweigh the benefits of multiple recombinant possibilities for organizations in different industrial domains.
- Collaboration: Joint patent development positively affects influence within biotechnology but shows no significant effect across the other domains.About 90% of joint patents involve biotechnology firms or biotechnology–pharmaceutical collaborations, which may keep their influence within biotechnology.
- Claims: The number of claims has an inverted-U relationship with influence in most domains but a linear positive relationship outside biotechnology.More claimed functionalities increase uncertainty about the patent’s application domain, potentially reducing its value as a source for subsequent technology.
- Scope: Scope has a curvilinear effect outside the assignee’s organization but a linear positive effect outside biotechnology.Broad industrial applicability may support external applications, whereas excessive scope may reflect strategic choices and reduce value for other organizations.
- Novelty: Novelty has a curvilinear relationship with overall influence and influence across industrial and organizational boundaries.Very high novelty may slow diffusion because firms outside biotechnology can have difficulty absorbing and using the invention.
6. CONCLUDING REMARKS
The paper examines how six patent determinants shape influence differently across industrial and organizational domains, using a multidimensional analysis of biotechnology patents. It finds that patenting strategies and resource allocation should reflect the specific type of impact sought.
- Method: 5671 biotechnology patents from 1976 to 2003 underpin the empirical analysis of four impact domains.
- Contribution: The study analyzes six patent determinants across influence within and outside industrial and organizational boundaries.These determinants include scientific knowledge, technological breadth, collaboration, claims, scope, and novelty.
- Findings: Patent influence varies across industrial and organizational domains, so determinants do not have uniform effects across impact types.
- Findings: Joint development strengthens technological relevance within biotechnology, whereas broader scope strengthens influence outside biotechnology.
- Implications: Scientific knowledge supports subsequent technologies of the assignee and motivates firms to strengthen their scientific base through recruitment and research collaborations.
- Limitations: The analysis is limited by citation-based influence measures, unexamined interdependencies among indicators, and a sample restricted to US biotechnology firms.