Krugman suggests that there is a convincing economic rationale behind these concerns
which can be illustrated by the example of externalities arising from knowledge spillovers.
While there exist other possible causes of technological and linkage externalities, the generation of knowledge is a particularly useful target because it is “an activity unlikely to be rewarded at its marginal social productivity.” Patents and other intellectual property rights are meant to ensure that the developers of knowledge benefit from their ideas at least for a time, but such instruments are by no means perfect and cannot completely prevent what one firm learns through experience or R&D from being appropriated by another firm.15 Because the social benefits of the knowledge generated are not fully represented in private returns, there is a role for the government to provide subsidies that encourage R&D. Carried to its full conclusion, this would suggest that the best policy is to subsidize all firms that engage in the generation of knowledge in order to ensure that the socially optimal level of research is achieved and national welfare is maximized. Furthermore, if foreign promotion of such knowledge-generating sectors is occurring, then subsidization appears to be particularly crucial as the home country may be deprived of valuable spillovers if the government fails to counteract the advantages granted to foreign firms