Partial Communication in a Voluntary‐Contribution‐Mechanism Experiment

This paper presents results of a voluntary‐contribution‐mechanism experiment with partial communication. It was found that the trend of mean investments depended on whether a communication network was connected and on how fast all subjects could share information. When each subject could communicate with two other subjects and the network was connected, the mean investment increased from 33 percent to 58 percent of the optimal investment. On the other hand, when each subject could communicate with only one other subject and the network was not connected, the mean investment decreased from 34 percent to 19 percent of the optimal investment.     

Contributions and Channels of Interindustry R&D Spillovers: An Estimation for Japanese High-tech Industries

This paper estimates the contributions of R&D spillovers to four high-tech industries in Japan—general machinery, electrical machinery, transportation machinery and chemicals—by estimating the trans-log cost and share functions that include the R&D stock variables of own and spillover-source industries. The candidates for spillover- source industries are selected on the basis of large R&D flow or R&D proximity. The R&D flow measures the spillover embodied in purchased intermediate goods using input–output coefficients. The R&D proximity measures the extent of similarity between a pair of industries of the distribution of R&D expenditures across research fields, and is expected to show the likelihood of spillover at the R&D stage. The results suggest that electrical machinery benefited from R&D in the chemical industry, through the purchase of intermediate goods, whereas general machinery and transportation machinery benefited from R&D in the metal products industry, through R&D proximity. There was no evidence of the chemical industry benefiting from R&D spillovers. These results clearly imply that the contributions and the channels of R&D spillovers are diverse, casting doubt on earlier studies that used weighted sums of R&D expenditures (or their stocks) of other industries as aggregate spillover variables.