Trademarks Statistics as Innovation Indicator? - A Micro Study

This study explores the possibility of using trademark statistics as an indicator of new-to-the-firm innovation. It compares registrations of trademarks and the launch of new-to-the-firm products for a number of Swedish companies in the electromechanical, automotive and pharmaceutical industries. The results show considerable differences between these sectors. For the studied companies in the electromechanical and automotive industries, trademarks are generally unreliable as an indicator of new products. However, the results from the pharmaceutical industry are different. A comparison between trademarks and the number of new products show that a high and fairly stable percentage of the new products have been trademarked. The number of trademarks is also mutually co-integrated with the number of registered (approved) new drugs over the time period 1935-1996. Short term variations are not correlated, possibly due to the extreme uncertainty of the development process in the pharmaceutical industry, where many projects fail even in the final testing, as well as the often lengthy registration process. A comparison between the number of trademarks and patents nevertheless suggests that the trademark statistics carries information about product development activities prior to registration. The study also suggests that the use of trademarks as an indicator of new-to-the-firm innovation is likely to be most successful in industries with frequent use of trademarks and with products targeting consumers or professional end-users, but with less uncertain development process than the pharmaceutical industry’s.