| dc.description.popularabstract | This thesis investigates the willingness of firms to cooperate on Research and Development (R&D). I present three papers concerning incentives to share/exchange knowledge created by the firms’ own R&D. It is my hope that this thesis can give some clues of how sharing and exchange of knowledge can contribute to an already large literature on Research and Development.
It is commonly believed that in order to sustain future growth and secure increased living standards, R&D is one of the few and most important ways to achieve this goal. R&D conducted by firms and universities create a lot of knowledge which has the possibility to be turned into innovation. Innovation creates new products and processes, products that replaces old ones and processes that increase efficiency and reduce cost. As a testimony to the perceived importance of R&D, the European Union has set a goal that a certain fraction of each country’s GDP should be committed to R&D. Public polices have also been changed to foster more R&D. For example, the National Cooperation Act was enacted in the U.S in 1984, with the aim of fostering R&D by allowing firms to cooperate on R&D. Prior to this Act, firms were not allowed to conduct R&D together, because it was feared to lead to collusion in the product market.
To investigate this topic, this thesis employs economic laboratory experiments. A typical economic Industrial Organization (IO) laboratory experiment has subjects conduct choices in a computer lab, where each participant takes on the role as one firm. An important role of IO experiments is to bridge theoretical modeling and empirical work done on real life data. Some of the benefits of using such experiments include control over the environment within which the firms interact. Collecting data from real firms have several issues that make inference problematic. A major problem is the potential for reverse causality. Does competition affect innovation or does innovation affect competition? Another issue is related to measurement. For example, how can one accurately measure competition? A laboratory experiment tries to address some of these issues by allowing for human behavior rather than economic man behavior. Experimenters may control the information and parameters as they wish. In principle, varying one variable between different treatments, a change in the outcome variable they are measuring should be attributed to the variable the experimenter is manipulating. In other words, a causal effect. However, laboratory experiments do come with some drawbacks. Since mostly students are recruited as participants, the field is criticized for lack of representativeness. That is, how can public policy be based on the decision of students sitting in front of a computer? Experiments does not paint the whole picture. They can however, provide evidence and valuable links for the theoretical modeling and empirical work based on economics and field data.
Let me summarize some of the main findings of this thesis. I find that firms are less willing to cooperate on R&D when the firms face tough market competition. It is not just competition that plays a role but also initial asymmetry between firms. If one firm is ahead of the others in terms of technology or knowledge, the leading firm will be less willing to agree to cooperate on R&D. Allowing for a side payment being transferred from the lagging firm to the leading one makes the leading firm more willing to cooperate. One way that seems to lead to both high R&D and incentive to cooperate, is to prevent firms from entering into R&D exchange before investment in R&D has taken place. The threat of one firm “out running” the other, keep both investing, and as long as the firms invested fairly equally, they wished to cooperate by agreeing ex post to share their knowledge. | en_US |
English
norsk