By reference to the R&D project management of Japanese high-tech firms in the 1980s, this paper demonstrates a stable multiplier of 1.25 between R&D expenditures and capital investments. Given the assumption that the amount of plant and equipment investment represents a firm's subjective estimate of the expected market, the inference is that the Japanese firms preferred to invest in R&D projects which could guarantee multiplier effects on the size of the market, so that the successful projects showed economies of scale in their R&D expenditure. The analysis leads us to assume that Japanese firms in the 1980s preferred to invest in projects conditioned by a fixed manageable risk, investing cumulatively in the revealed direction of technical change. With the demise of demand expansion in the 1980s, however, bringing out 'extraordinary innovations has become the R&D target of many Japanese firms. This new target requires the firms to accept a wider R&D agenda and longer R&D lead times and to follow the (possibly somewhat paradoxical) guideline of not sticking at manageable risk.
