Abstract
This paper examines the generation of technological knowledge by leading companies in the defence industry. In particular, we test whether the characteristics of large defence companies are related to both the production of different types of patents (civilian, military and mixed), and the generation of dual-use technologies. To explore these links, we rely on economic data for the top 100 defence companies from the Stockholm International Peace Research Institute database, and patent information from the Worldwide Patent Statistical Database. Our results show that the relationship between the production of civilian patents and the size of the company is positive and significant. However, this relationship does not hold for the production of military patents. Furthermore, the military commercial profile is unrelated to the generation of military patents. Regarding the involvement in dual-use technologies, firms engaged in dual-use are those with higher military sales, a greater number of employees and a larger number of patents (civilian, military and mixed) than those not engaged in dual-use. Furthermore, we found a skill effect (more involvement in dual-use per employee) in European firms compared to US firms. These findings help to identify which firms should be targeted by government policies if increasing dual-use technologies becomes a political objective.
Acknowledgements
The authors are also very grateful to an anonymous reviewer for constructive and insightful comments on an earlier version of this paper.
Notes
1. These firms account for a large part of the defence industry. For example, in the case of Europe, the study for the European Commission by Cauzic et al. (Citation2009) estimated the direct employment in the European defence industry as about 1,172,200 employees (including not just the production of weapons, but a wide range of activities), and 31.75% of this figure was provided by prime contractors.
2. The SIPRI website (www.sipri.org) provides detailed information on the data-set and the definitions of the variables.
3. We also eliminated one firm for which the employment was 0, and another firm that did not have data for total sales.