Incentive Pay or Windfalls: Remuneration for employee inventions in Japan

This paper summarizes historical developments in Japan's legal treatment of firms' invention remuneration policies and examines the impact of such policies on R&D performance using original data from surveys including the 2005 IIP Invention Remuneration Survey, the 2007 RIETI Inventor Survey and its 2008 follow-up survey. Tracking the linkages between remuneration policy and R&D performance is complicated by Japanese firms' reluctance to reveal the details of their policies to their employees before the 2004 amendment of Japan's Patent Law. By matching the data from firm-level and individual-level surveys, we find that nearly 40% of inventors believed that their firms did not have revenue-based remuneration although their employers reported they actually had instituted such policies. We estimate the effect of revenue-based remuneration policies on R&D performance using two policy variables for the incidence of contingent remuneration policies, one of which depends on the firms' responses and the other on individual employees' survey responses. Our project-level or patent-level productivity measures include the number of patents generated from a project, inventors' subjective evaluation of their patents' economic value, and indicators of whether the patent was used commercially. After taking measures to eliminate estimation bias, such as the propensity score matching method and instrumental variable estimation, we find no evidence that the prospect of greater monetary compensation is affecting the quantity or quality of patents or the likelihood of their commercialization except for a sample of those who rated the importance of monetary rewards as "not unimportant", from whom we obtain weak evidence of the real impact of monetary incentives.