Experimental studies on strategic research and development
von Andreas NicklischThe issue of persistent economic change appears to be very important for economics if we understand competition as an inherently dynamic process where markets continuously evolve. There is a rich theoretical (e. g., Dasgupta & Stiglitz, 1980, Harris & Vickers, 1987, Kamien, Muller & Zang, 1992) and empirical literature (e. g., Lerner, 1997, Cassiman & Veugelers, 2002) extending this idea. Yet experimental evidence on this topic is rare (e. g., Isaac & Reynolds, 1988, Zizzo, 2002). Of course, laboratory experiments cannot provide a framework in which actual invention processes are observed. Nevertheless, experimental studies illustrate important aspects of this problem as they focus on the strategic interaction between agents engaged in activities on research and development (hereafter R& D). Most prominently, the influence of market concentration on the level of research expenditures is analyzed (e. g., Isaac & Reynolds, 1992). This thesis attempts to provide a further step in exploring individual research behavior. The studies that are outlined in the three chapters analyze the process of investment decisions for search activities. Thereby, this study attempts to advance the understanding of factors that control search activities by observing behavioral regularities in experimental settings. The main experimental findings indicate that competitors in this context base their decision whether to increase or decrease research effort on comparison of relative performances.
In order to explore search investments systematically, we introduce in a first study a prototype experiment that allows an investigation of individual R& D decisions. Of course, in a laboratory setting we do not observe actual invention processes but processes of strategic interaction on research decisions. Firms attempt to improve their products by varying some or all aspects of their products. Search investments subsequently reveal ideal aspects which yield payoffs. The effects of information spillovers result from the imperfect appropriability of research successes. We identify the relative performance of firms (measured as the relative distance between accumulated payoffs) as the crucial factor in determining the timing of innovative attempts. If competitors possess a rather homogenous economic wealth, we observe a significantly higher probability of further search investments than if competitors maintain a large difference in their economic wealth.
The second study applies the idea that relative performance promotes search activities to sequential patent races. Therefore we extend the prototype game for individual R& D decisions by an exogenous shock which affects both the supply and demand side and again spurs research competition. Consequently, after the exogenous shock, heterogeneous firms – a well established market participant and a less successful, economically lagging market participant – invest in search activities. Thus, the experiment allows for an analysis that explores the relation between market concentration and innovative activities. Again, we find that relative economic performance exercises a crucially stimulating effect on the implementation of innovation and imitation strategies. Particularly, we can differentiate among different types of innovation. Homogeneous economic wealth promotes intensified search investments for incremental innovation. However, we observe more investment for fundamental innovation if competitors possess a rather heterogeneous economic wealth.
Finally, a third study investigates the relationship between innovation competition and price competition. Here, firms are engaged in Bertrand competition for heterogeneous goods. Additionally, similar to the study by Nagel & Vriend (1999), they can invest in product quality. The investments affect their payoffs by increasing the maximal demand of the product. Yet player’s investment shifts not only the own demand function but also the opponent’s demand function. Thus, the imperfect approriability of investments creates spillover effects between firms. The experimental results allow us to analyze the individual adaptation processes in this complex environment. Again, we find that the relative competitiveness exercises an important influence on investment decisions. A poor relative performance leads to a change in investments. Moreover, we can formalize the relation between relative competitiveness and behavioural change in a model of interactive learning which anticipates competitors’ response on differences in relative performances. This approach of social learning is rarely recognized in the economic literature on learning (e. g., see Huck, Norman & Oechssler, 2004).
Analyzing the relevance of behavioural principles is fundamental to understand market dynamics; therefore, the experimental approach has been pursued. Of course, this thesis provides only a starting point for the assessment of individual search decisions. In particular, further studies have to generalize the approach of social learning. Nonetheless, the results of this thesis indicate that relative competitiveness is a crucial factor for decisions in complex environments.
In order to explore search investments systematically, we introduce in a first study a prototype experiment that allows an investigation of individual R& D decisions. Of course, in a laboratory setting we do not observe actual invention processes but processes of strategic interaction on research decisions. Firms attempt to improve their products by varying some or all aspects of their products. Search investments subsequently reveal ideal aspects which yield payoffs. The effects of information spillovers result from the imperfect appropriability of research successes. We identify the relative performance of firms (measured as the relative distance between accumulated payoffs) as the crucial factor in determining the timing of innovative attempts. If competitors possess a rather homogenous economic wealth, we observe a significantly higher probability of further search investments than if competitors maintain a large difference in their economic wealth.
The second study applies the idea that relative performance promotes search activities to sequential patent races. Therefore we extend the prototype game for individual R& D decisions by an exogenous shock which affects both the supply and demand side and again spurs research competition. Consequently, after the exogenous shock, heterogeneous firms – a well established market participant and a less successful, economically lagging market participant – invest in search activities. Thus, the experiment allows for an analysis that explores the relation between market concentration and innovative activities. Again, we find that relative economic performance exercises a crucially stimulating effect on the implementation of innovation and imitation strategies. Particularly, we can differentiate among different types of innovation. Homogeneous economic wealth promotes intensified search investments for incremental innovation. However, we observe more investment for fundamental innovation if competitors possess a rather heterogeneous economic wealth.
Finally, a third study investigates the relationship between innovation competition and price competition. Here, firms are engaged in Bertrand competition for heterogeneous goods. Additionally, similar to the study by Nagel & Vriend (1999), they can invest in product quality. The investments affect their payoffs by increasing the maximal demand of the product. Yet player’s investment shifts not only the own demand function but also the opponent’s demand function. Thus, the imperfect approriability of investments creates spillover effects between firms. The experimental results allow us to analyze the individual adaptation processes in this complex environment. Again, we find that the relative competitiveness exercises an important influence on investment decisions. A poor relative performance leads to a change in investments. Moreover, we can formalize the relation between relative competitiveness and behavioural change in a model of interactive learning which anticipates competitors’ response on differences in relative performances. This approach of social learning is rarely recognized in the economic literature on learning (e. g., see Huck, Norman & Oechssler, 2004).
Analyzing the relevance of behavioural principles is fundamental to understand market dynamics; therefore, the experimental approach has been pursued. Of course, this thesis provides only a starting point for the assessment of individual search decisions. In particular, further studies have to generalize the approach of social learning. Nonetheless, the results of this thesis indicate that relative competitiveness is a crucial factor for decisions in complex environments.