A Duopoly Experiment on Cooperative and Noncooperative R&D

In this paper an experimental environment to test theoretical predictions concerning R&D behavior of firms in duopoly with allowance for R&D spillovers is created. The design and hypotheses of the experiment are based on the well-known model of d Aspremont and Jacquemin in which R&D behavior of firms either competing or cooperating in R&D, is calculated. No di.erence in behavior between di.erent spillover levels is found. Further, I find that irrespective of the technological spillover level subjects do not always commit to an R&D contract but if they commit to an R&D level in a binding contract, cooperative R&D levels are chosen. When subjects do not or cannot commit to a contract, the subgame perfect Nash equilibrium performs well in predicting R&D decisions.