In this paper we consider the case for subsidies towards firms which genera
te R&D spillovers in open economies. We show that in the presence of strate
gic behaviour by firms many expected results are overturned. Local R&D spil
lovers to other domestic firms may justify an R&D tax rather than a subsidy
; R&D cooperation by local firms over-internalises the externality and also
justifies an R&D tax; and international spillovers which benefit foreign f
irms may justify a subsidy, even though the government cares only about the
profits of home firms.